[go: up one dir, main page]

CN117422716B - Ecological early warning method and system for broiler chicken breeding based on artificial intelligence - Google Patents

Ecological early warning method and system for broiler chicken breeding based on artificial intelligence Download PDF

Info

Publication number
CN117422716B
CN117422716B CN202311743590.2A CN202311743590A CN117422716B CN 117422716 B CN117422716 B CN 117422716B CN 202311743590 A CN202311743590 A CN 202311743590A CN 117422716 B CN117422716 B CN 117422716B
Authority
CN
China
Prior art keywords
cockscomb
crown
area
trend
tips
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202311743590.2A
Other languages
Chinese (zh)
Other versions
CN117422716A (en
Inventor
魏叶堂
武军
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yishui Youbang Breeding Service Co ltd
Original Assignee
Yishui Youbang Breeding Service Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Yishui Youbang Breeding Service Co ltd filed Critical Yishui Youbang Breeding Service Co ltd
Priority to CN202311743590.2A priority Critical patent/CN117422716B/en
Publication of CN117422716A publication Critical patent/CN117422716A/en
Application granted granted Critical
Publication of CN117422716B publication Critical patent/CN117422716B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/02—Agriculture; Fishing; Forestry; Mining
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H30/00—ICT specially adapted for the handling or processing of medical images
    • G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10024—Color image

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Public Health (AREA)
  • Primary Health Care (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Radiology & Medical Imaging (AREA)
  • Business, Economics & Management (AREA)
  • Epidemiology (AREA)
  • Databases & Information Systems (AREA)
  • Mining & Mineral Resources (AREA)
  • Quality & Reliability (AREA)
  • Data Mining & Analysis (AREA)
  • Biomedical Technology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Agronomy & Crop Science (AREA)
  • Animal Husbandry (AREA)
  • Marine Sciences & Fisheries (AREA)
  • Pathology (AREA)
  • Geometry (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Image Processing (AREA)

Abstract

The invention relates to the technical field of image analysis, in particular to an artificial intelligence-based broiler chicken raising ecological early warning method and system. Firstly, dividing a to-be-determined cockscomb area in a monitoring image of broiler chickens; according to the direction trend situation of the pixel points at the edge of the cockscomb to be determined in the cockscomb to be determined area, determining trend pixel points from the pixel points at the edge of the cockscomb to be determined; connecting trend pixel points to obtain trend segments, and determining a target cockscomb area based on the number of the trend segments in the cockscomb area to be determined; determining the crown apex in the target cockscomb area; determining the crown tips belonging to the same broiler chicken according to the peripheral gray level distribution of the crown tips in the target crown region and the inclination angle change of the crown tips; detecting the crown tips of each broiler chicken in the target comb region to obtain the health state of the broiler chicken, and carrying out early warning. Under the condition that the raising density of the broiler chickens in the raising farm is high, the cockscomb belonging to the same broiler chickens is determined through the analysis of the cockscombs, so that the accurate detection of the health state of the broiler chickens is realized.

Description

Ecological early warning method and system for broiler chicken breeding based on artificial intelligence
Technical Field
The invention relates to the technical field of image analysis, in particular to an artificial intelligence-based broiler chicken raising ecological early warning method and system.
Background
In large-scale broiler breeding, the number of the groups of broiler chickens is large, and the density of broiler chicken breeding is relatively high, so that serious consequences can be brought to broiler chicken breeding when infectious diseases occur in the broiler chicken breeding process. Therefore, the health state of the broiler chickens needs to be detected, and the sick chickens in the broiler chickens can be found in time. Because the health state of the chickens determines the color of the cockscomb, the chicken is usually indicated to be possibly infected by infectious diseases such as avian influenza when the cockscomb is dark red; when the cockscomb is purple black, the possible pesticide poisoning of the broiler chicken is indicated; when the cockscomb is blue-purple, the chicken may suffer from some toxic diseases. Therefore, the health state of the broiler chickens can be judged through the color of the cockscomb. The cockscomb color of each chicken in the farm is extracted, which is critical for the health status detection of broilers.
The common method for detecting the color of the cockscomb of the broiler chicken at present is to acquire the cockscomb of the broiler chicken in an image through threshold segmentation by utilizing the characteristic that the cockscomb of the broiler chicken has obvious color difference with other parts of the broiler chicken. However, since the raising density of the farm is generally high, a cocktail containing a plurality of broilers in one communicating area may be caused, and the cocktails of the broilers may overlap in the acquired images, so that it is difficult to identify the number of sick chickens in the farm by identifying the health condition of the cocktail, and it is difficult to identify the health status of each broiler.
Disclosure of Invention
In order to solve the technical problem that the health state of each broiler chicken is difficult to identify due to the fact that cockscombs of the broiler chicken possibly overlap in the acquired images, the invention aims to provide an artificial intelligence-based ecological early warning method and system for broiler chicken breeding, and the adopted technical scheme is as follows:
in a first aspect, an embodiment of the present invention provides an artificial intelligence-based ecological early warning method for broiler chicken breeding, which includes the following steps:
acquiring a monitoring image of broiler chickens; dividing a pending cockscomb area of the broiler monitoring image;
determining trend consistency of the pixel points at the edge of each pending cockscomb according to the direction trend situation of the pixel points at the edge of the pending cockscomb in the pending cockscomb area; determining trend pixel points from the pixel points at the edge of the cockscomb to be determined according to the trend consistency; connecting trend pixel points to obtain trend segments, and determining a target cockscomb area based on the number of the trend segments in the cockscomb area to be determined;
a line segment connecting the target cockscomb area, and determining a cocktail; determining the crown tips belonging to the same broiler chicken according to the peripheral gray level distribution of the crown tips in the target crown region and the inclination angle change of the crown tips;
detecting the crown tips of each broiler chicken in the target comb region to obtain the health state of the broiler chicken, and carrying out early warning according to the health state of the broiler chicken.
Preferably, the determining the trend consistency of the pixel points at the edge of the cockscomb according to the trend situation of the pixel points at the edge of the cockscomb in the cockscomb area includes:
acquiring gradient direction values corresponding to pixel points at the edge of each pending cockscomb in the pending cockscomb area;
taking the left-most cockscomb edge pixel point in each cockscomb area as a chain code starting point corresponding to each cockscomb area, and constructing an edge direction chain code corresponding to the cockscomb area by using a gradient direction value corresponding to each cockscomb edge pixel point;
and taking any element in the edge direction chain code as a target element, and taking the number of elements which are the same as the numerical value of the target element in a sliding window corresponding to the target element as the trend consistency of the pixel points of the edge of the cockscomb to be determined corresponding to the target element.
Preferably, the determining the crown point belonging to the same broiler chicken according to the peripheral gray level distribution of the crown point in the target comb region and the inclination angle change of the crown point comprises:
calculating the probability that the crown tips and the crown tips at two sides belong to the same broiler chicken according to the peripheral gray distribution of the crown tips in the target crown region and the change of the inclination angle of the crown tips;
when the probability that the crown tips and the crown tips at two sides belong to the same broiler chicken is larger than or equal to a preset probability threshold value, judging that the crown tips and the crown tips at two sides corresponding to the crown tips belong to the same broiler chicken.
Preferably, the calculating the probability that the crown point and the crown points at two sides belong to the same broiler chicken according to the peripheral gray level distribution of the crown point in the target comb region and the inclination angle change of the crown point includes:
the calculation formula of the probability that the crown tips and the crown tips at two sides belong to the same broiler chicken is as follows:
the method comprises the steps of carrying out a first treatment on the surface of the Wherein (1)>The probability that the (d) th crown apex, the (d-u) th crown apex and the (d+v) th crown apex belong to the same broiler chicken in the (R) th target crown area; />Is a sighard function; min is a function taking the minimum value; max is a maximum function; />The average gray value of each pixel point in the closed area corresponding to the d-th crown peak is obtained; />The average gray value of each pixel point in the closed area corresponding to the d-u crown points; />The average gray value of each pixel point in the closed area corresponding to the d+v crown points;is the inclination angle of the (d+v) th crown tip; />Is the inclination angle of the d-th crown tip; />Is the inclination angle of the d-u crown tip; />As a sign function.
Preferably, the determining the target cockscomb area based on the number of trend segments in the cockscomb area includes:
determining the smoothness of the region according to the number of trend segments in the cockscomb region to be determined;
and when the region smoothness is smaller than a preset smoothness threshold, taking the undetermined cockscomb region corresponding to the region smoothness as a target cockscomb region.
Preferably, the determining the smoothness of the region according to the number of trend segments in the cockscomb region comprises:
and taking the ratio of the area of the cockscomb area to the number of trend segments in the cockscomb area as the area smoothness of the cockscomb area.
Preferably, the determining the crown apex by connecting the line segments in the target comb region includes:
connecting line segments in the target cockscomb area to obtain at least two undetermined angles;
and when the undetermined angle is an acute angle and the inner area of the undetermined angle belongs to the target cockscomb area, taking the undetermined angle as a cockscomb tip.
Preferably, the segmenting the pending cockscomb area of the monitoring image of the broiler chicken includes:
and carrying out threshold segmentation on the broiler monitoring image, and marking the region obtained by threshold segmentation as a cocked comb region.
Preferably, the determining the trend pixel point from the pixel points at the edge of the cockscomb according to the trend consistency includes:
and taking the undetermined cockscomb edge pixel points with trend consistency larger than or equal to a preset trend threshold value as trend pixel points.
In a second aspect, an embodiment of the present invention provides an artificial intelligence-based ecological early warning system for broiler chicken raising, which includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor, where the processor implements the artificial intelligence-based ecological early warning method when executing the computer program.
The embodiment of the invention has at least the following beneficial effects:
firstly, dividing a to-be-determined cockscomb area in a monitoring image of broiler chickens; according to the direction trend situation of the pixel points at the edge of the cockscomb to be determined, determining trend pixel points from the pixel points at the edge of the cockscomb to be determined, wherein the trend pixel points reflect the trend change situation of the edge of the cockscomb to be determined, and the trend pixel points are combined to judge the target cockscomb area in the cockscomb to be determined because the cockscomb to be determined contains the beard area of the non-cockscomb area and has more change trend for the beard area; connecting trend pixel points to obtain trend segments, and determining a target cockscomb area based on the number of the trend segments in the cockscomb area to be determined, wherein the target cockscomb area is a cockscomb area excluding a beard area; connecting line segments in a target cockscomb area, and determining a cocktail; according to the peripheral gray level distribution of the crown tips in the target crown region and the inclination angle change of the crown tips, the crown tips belonging to the same broiler chicken are determined, and under the condition that the raising density of the broiler chicken in a raising farm is high, the condition that the crowns of a plurality of broiler chickens overlap can exist in a large probability, and under the condition that the crowns of more than one broiler chicken exist in the target crown region, the crown tips in the target crown region are analyzed, and the crown tips belonging to the same broiler chicken are determined; and detecting the crown tips of each broiler chicken in the target comb region to obtain the health state of the broiler chicken. According to the invention, under the condition that the raising density of the broiler chickens in the raising farm is high, the cockscomb belonging to the same broiler chickens is determined by analyzing the cockscomb, so that the accurate detection of the health state of the broiler chickens is realized.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flow chart of a broiler raising ecological early warning method based on artificial intelligence according to an embodiment of the invention.
Detailed Description
In order to further explain the technical means and effects adopted by the invention to achieve the preset aim, the following detailed description is given below of the artificial intelligence-based broiler raising ecological early warning method and system according to the invention, which are specific implementation, structure, characteristics and effects thereof, with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The embodiment of the invention provides a broiler raising ecological early warning method and a specific implementation method of a broiler raising ecological early warning system based on artificial intelligence. A monitoring device is arranged in the farm under the scene so as to photograph the broiler chickens in the farm, and a broiler chicken monitoring image is obtained. In order to solve the technical problem that the cockscomb of the broiler chickens can overlap in the acquired images, so that the health state of each broiler chicken is difficult to identify. The method analyzes the crown tips in the target cockscomb area and determines the crown tips belonging to the same broiler chicken; and detecting the crown tips of each broiler chicken in the target comb region to obtain the health state of the broiler chicken. Under the condition that the raising density of the broiler chickens in the raising farm is high, the cockscomb belonging to the same broiler chickens is determined by analyzing the cockscomb, so that the accurate detection of the health state of the broiler chickens is realized.
The invention provides a broiler raising ecological early warning method and a broiler raising ecological early warning system based on artificial intelligence, which are concretely described below with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of steps of an artificial intelligence-based ecological early warning method for broiler chicken raising according to an embodiment of the invention is shown, and the method comprises the following steps:
step S100, acquiring a monitoring image of broiler chickens; and dividing the undetermined cockscomb area of the broiler monitoring image.
And shooting the broiler chickens in the farm through monitoring equipment in the farm so as to obtain a broiler chicken monitoring image. It should be noted that the monitoring image of the broiler chicken is an RGB image. Searching images of various sick chickens and normal chickens on the Internet to obtain an image set of the sick chickens and the normal chickens。
And analyzing the cockscomb areas of the image concentrated sick chickens and the normal chickens to obtain color intervals of cockscombs of the sick chickens and cockscombs of the normal chickens under each RGB channel under the RGB image. Wherein, the cockscomb of the sick chicken and the cockscomb of the normal chicken are collectively called as cockscomb of the chicken. And calculating the average value of the cocks of the chickens in the image set in each channel of RGB, and recording the average value as a preset cockscomb channel value.
Because the cocks and the beards of the chickens have obvious color differences with other parts of the chickens, each channel based on RGB is used, the region of the cocks and the beards is obtained from the monitoring images of the chickens by threshold segmentation, the region is marked as a cocked cockscomb region, namely the monitoring images of the chickens are subjected to threshold segmentation, and the region with small difference with the preset cockscomb channel value in the two regions obtained by threshold segmentation is marked as a cocked cockscomb region.
In another embodiment of the present invention, the obtained monitoring images of the broiler chickens and the images in the image set are gray images; and acquiring a gray average value of the cockscombs of the chickens in the image set, and marking the gray average value as a preset cockscomb channel value. And (3) carrying out threshold segmentation on the broiler monitoring image, and marking the region with smaller difference from the preset cockscomb channel value in the two regions segmented by the threshold as a cockscomb region. It should be noted that there is more than one cockscomb area to be identified.
It should be noted that, the region of the comb and the beard in the monitoring image of the broiler chicken is divided, but not the region of the comb is divided, because the colors of the comb and the beard are relatively close, and the region of the comb and the beard is difficult to be distinguished only by threshold segmentation. So in the embodiment of the invention, firstly, the cocked comb region containing the comb and the beard is segmented; further, the characteristics of the beards and the cocks are analyzed to determine a target cockscomb area corresponding to the real cockscomb in the cockscomb area to be determined.
Step S200, determining trend consistency of the pixel points at the edge of the cockscomb to be determined according to the direction trend situation of the pixel points at the edge of the cockscomb to be determined in the cockscomb to be determined; determining trend pixel points from the pixel points at the edge of the cockscomb to be determined according to the trend consistency; and connecting the trend pixel points to obtain trend segments, and determining a target cockscomb area based on the number of the trend segments in the cockscomb area.
After the cocked comb area containing cocks and beards is obtained according to the step S100, further, a target comb area corresponding to the real cocks is determined from the cocked comb area.
Because the cocktail of the broiler chicken is provided with cocktail tips, more small trend segments exist on the edge of the region where the cocktail is located. The beards are more round than the cockscomb, so that more small trend segments are not arranged on the communicating domain where the beards are positioned. The target cockscomb area in the cockscomb area is obtained through the size of each connected domain and the number of trend segments contained in the edge of the connected domain.
In the embodiment of the invention, the morphological characteristics of the region are described through the chain codes, the directions of the subsequent pixel points relative to the current pixel point are recorded, and the opposite vectors are converted into numerical values. A trend segment is represented in the chain code as a segment of the same value of the chain code. The number of trend segments contained in the edge of each region is obtained by the chain code of the edge line of each region.
Firstly, gradient direction values corresponding to pixel points at the edge of each pending cockscomb in the pending cockscomb area are obtained. It should be noted that, the pixel points at the edge of the cockscomb to be detected are the pixel points on the edge line of the cockscomb to be detected.
And taking the left-most pending cockscomb edge pixel point in each pending cockscomb area as a chain code starting point corresponding to each pending cockscomb area. If a plurality of pixel points at the edge of the cocks are arranged at the leftmost side, the pixel point at the edge of the cocks at the leftmost side is used as a chain code starting point corresponding to the cocks area.
And constructing an edge direction chain code corresponding to the cockscomb area to be determined according to the gradient direction value corresponding to each cockscomb edge pixel point to be determined. It should be noted that, the gradient direction value corresponding to each pixel point at the edge of the cockscomb to be determined is an angle value formed by the gradient direction corresponding to each pixel point at the edge of the cockscomb to be determined and a horizontal line to the right.
And taking any element in the edge direction chain code as a target element, and taking the number of elements which are the same as the numerical value of the target element in a sliding window corresponding to the target element as the trend consistency of the pixel points of the edge of the cockscomb to be determined corresponding to the target element. In the embodiment of the invention, the size of the sliding window is n×1, and the value of n is 7.
When the chain codeMiddle->Individual chain code->The window has a plurality of values of chain code elements and +.>When the values of the chain code elements are the same, the more the number of the elements which are the same as the values of the target elements in the sliding window is, the value of the trend consistency is close to 1, which indicates the chain code +.>Middle->The chain code has stronger consistency with the surrounding chain codes, i.e. the pixel point represented by the chain code is more likelyThe pixel point in a data trend segment; when the chain code->Middle->Individual chain code->The window has the value of less chain code elements and +.>When the values of the chain code elements are the same, the number of the elements which are the same as the value of the target element in the sliding window is smaller, and the value of the trend consistency approaches 0 at the moment, which indicates the chain code +.>Middle->The individual chain codes have no strong consistency with the surrounding chain codes, i.e. the pixel points represented by the chain codes are more likely not to be the pixel points in a data trend segment.
Further, according to trend consistency, determining trend pixel points from the pixel points at the edge of the cockscomb to be determined, and specifically: and taking the undetermined cockscomb edge pixel points with trend consistency larger than or equal to a preset trend threshold value as trend pixel points. In the embodiment of the present invention, the preset trend threshold value is 0.7, and in other embodiments, the value can be adjusted by an implementer according to actual situations. When the trend consistency is greater than or equal to a preset trend threshold, reflecting that the pixel points at the edge of the cockscomb to be determined have stronger consistency with other pixel points in the corresponding edge direction chain code, and reflecting that the probability of the pixel points in the edge direction chain code corresponding to the pixel points at the edge of the cockscomb to be determined on a data trend is higher. When the trend consistency is smaller than a preset trend threshold, the consistency of the pixel points at the edge of the cockscomb to other pixel points in the corresponding edge direction chain code is reflected to be weaker, and the probability of the pixel points in the edge direction chain code corresponding to the pixel points at the edge of the cockscomb to be reflected to be on a data trend is smaller.
Performing the above operation on each chain code element in the edge direction chain code, namely performing the above operation on each pixel point at the edge of the cockscomb to obtain a trend pixel point set on the data trend segment in the cockscomb areaNamely, the trend pixel points are formed into a trend pixel point set. />Representing the +.sup.th in cockscomb area to be determined>And pixel points on the data trend segment.
Connecting the trend pixel points to obtain a trend segment, and specifically: and connecting adjacent trend pixel points to obtain a trend segment. I.e. a set of trend pixel pointsAnd connecting adjacent trend pixel points to obtain all trend segments in the undetermined cockscomb area.
Further, based on the number of trend segments in the cockscomb area, determining a target cockscomb area, specifically: determining the smoothness of the region according to the number of trend segments in the cockscomb region to be determined; and when the region smoothness is smaller than a preset smoothness threshold, taking the undetermined cockscomb region corresponding to the region smoothness as a target cockscomb region. In the embodiment of the present invention, the value of the preset smoothing threshold is 100, and in other embodiments, the value may be adjusted by an implementer according to the actual situation. When the region smoothness is larger than or equal to a preset smoothness threshold value, reflecting that the trend change in the pending cockscomb region is less, the probability that the corresponding pending cockscomb region is a bearded region is larger. When the region smoothness is smaller than a preset smoothness threshold value, reflecting that the trend in the cockscomb region is changed more, and the probability that the corresponding cockscomb region is a cockscomb region is larger.
The method comprises the steps of determining region smoothness according to the number of trend segments in a cockscomb region to be determined, and specifically: and taking the ratio of the area of the cockscomb area to the number of trend segments in the cockscomb area as the area smoothness of the cockscomb area.
The calculation formula of the region smoothness is as follows:
wherein,region smoothness for the i-th cockscomb region; />For the area of the i < th > cockscomb area, < in the embodiment of the present invention->Namely the number of pixel points in the ith cockscomb area; />Is the number of trend segments in the ith cocked comb area.
The number of trend segments in the cockscomb area reflects the number of trend segments contained in the cockscomb area, and the number of pixel points in the cockscomb area reflects the area of the cockscomb area. In the embodiment of the present invention, the number of pixel points in the region is used as the area size of the region, and in other embodiments, the area of the cockscomb region to be determined may be obtained by other methods.
When the trend segments contained in the cockscomb area are more and the area is smaller, namely the area of the cockscomb areaThe smaller the number of trend segments in the cockscomb area +.>The larger the area, the smaller the smoothness of the corresponding areaThe more likely the cockscomb area is the cockscomb area at this time, because the area of the cockscomb area is generally smaller than the area of the beard area, and the beard area is generally smooth and circular arc-shaped, and the cockscomb area is cockscomb-pointed, so the trend segment corresponding to the cockscomb area is more. Conversely, when the trend segment contained in the cockscomb area is smaller and the area is larger, i.e., the area of the cockscomb area +.>The greater the number of trend segments in the cockscomb area +.>The smaller the time, the larger the smoothness of the corresponding region, and the larger the probability that the pending cockscomb region is a bearded region.
Therefore, further, determining the target cockscomb area in the cockscomb area to be determined according to the area smoothness, and specifically: and when the region smoothness is smaller than a preset smoothness threshold, taking the undetermined cockscomb region corresponding to the region smoothness as a target cockscomb region. When the smoothness of the region is smaller than a preset smoothness threshold, reflecting that the trend in the cockscomb region to be determined is smaller and the area is larger; and when the region smoothness is greater than or equal to a preset smoothness threshold, the tendency in the pending cockscomb region is reflected to be larger, and the area is smaller. In the embodiment of the present invention, the value of the preset smoothing threshold is 100, and in other embodiments, the value may be adjusted by an implementer according to the actual situation.
And (3) through the characteristic of crown points on the cockscomb, carrying out the analysis on the cockscomb area to be determined, and determining target cockscomb areas in all cockscomb areas to be determined.
Step S300, connecting line segments in the target cockscomb area, and determining the cockscomb tip; and determining the crown tips belonging to the same broiler chicken according to the peripheral gray level distribution of the crown tips in the target crown region and the inclination angle change of the crown tips.
Due to the high breeding density of broiler chickens in a broiler farm, more than one comb may be present in a target comb area, but a plurality of combs may be present. Because the crown edge lines on each broiler comb are oblique and intersecting, the crown is determined by connecting line segments in the target comb area. Firstly, edge detection can be performed on the target cockscomb area to obtain line segments in a plurality of target cockscomb areas, and further, the line segments in the target cockscomb areas are connected to obtain at least two undetermined angles. And when the undetermined angle is an acute angle and the inner area of the undetermined angle belongs to the target cockscomb area, taking the undetermined angle as a cockscomb tip.
And taking the line segment corresponding to the crown tip as a crown tip edge line, and calculating the inclination angle of each crown tip edge line.
The calculation formula of the inclination angle is as follows:
wherein,is the edge line of the crown tip->Is a tilt angle of (2); tan is a tangent function; />Is the edge line of the crown tip->The abscissa of the leftmost pixel point; />Is the edge line of the crown tip->The abscissa of the rightmost pixel point; />Is the edge line of the crown apexThe ordinate of the leftmost pixel point; />Is the edge line of the crown tip->The ordinate of the rightmost pixel point in (a).
When there is more than one pixel on the rightmost side or leftmost side, the leftmost pixel or rightmost pixel is used as the leftmost pixel or rightmost pixel.
Because the raising density of the broiler chickens in the raising farm is possibly higher, the cocks of a plurality of broiler chickens can be in the same communicating domain, so that the health degree of the chickens corresponding to the communicating domain cannot be judged according to the cocks of the same chicken, the invention obtains the cocks belonging to the same broiler chickens according to the fact that the cocks of the same chicken are single-increment or single-decrement in the changing direction of the cocks, the average gray values of the pixel points in the cocks of the same chicken are relatively similar, the inclination angles of the first cocks and the second cocks are relatively similar, and the angles between the second cocks and the third cocks are relatively different.
According to the peripheral gray level distribution of the crown tips in the target cockscomb area and the change of the inclination angles of the crown tips, determining the crown tips belonging to the same broiler chicken, and specifically:
calculating the probability that the crown tips and the crown tips at two sides belong to the same broiler chicken according to the peripheral gray distribution of the crown tips in the target crown region and the change of the inclination angle of the crown tips;
when the probability that the crown tips and the crown tips at two sides belong to the same broiler chicken is larger than a preset probability threshold, judging that the crown tips and the crown tips at two sides corresponding to the crown tips belong to the same broiler chicken.
The calculation formula for calculating the probability that the crown tips and the crown tips at two sides belong to the same broiler chicken is as follows:
the method comprises the steps of carrying out a first treatment on the surface of the Wherein (1)>Is the (d) th target cockscomb areaProbability that the crown tip, the d-u crown tip and the d+v crown tip belong to the same broiler chicken; />Is a sighard function; min is a function taking the minimum value; max is a maximum function; />The average gray value of each pixel point in the closed area corresponding to the d-th crown peak is obtained; />The average gray value of each pixel point in the closed area corresponding to the d-u crown points; />The average gray value of each pixel point in the closed area corresponding to the d+v crown points; />Is the inclination angle of the (d+v) th crown tip; />Is the inclination angle of the d-th crown tip; />Is the inclination angle of the d-u crown tip;as a sign function.
Wherein,reflecting the gray value difference between the d-th crown tip and the adjacent d-u-th crown tip, when the gray values of the d-th crown tip and the d-u-th crown tip have no large difference, the +_s is changed by the +_s>The value of (2) is approximately 1, whereas +.>The value of (2) approaches 0; />Reflecting the gray value difference between the (d) th crown point and the adjacent (d+v) th crown point, when the gray values of the (d) th crown point and the (d+v) th crown point have no large difference, the (d+v) th crown point is characterized in that the (d) th crown point is a (d+v) th crown point>The value of (2) is approximately 1, whereas +.>The value of (2) approaches 0.
The angle of inclination of the d-th crown point is the angle between the left edge line of the d-th crown point and the horizontal line in the embodiment of the invention. />Reflecting whether the angle changes of the left edges of the three crown tips are the same or not, and when the angle changes of the left edges of the three crown tips are similar, corresponding +.>The closer to 1, the corresponding +.>The closer to 0.Reflects the similarity of the change of the inclination angles of the three crown tips, and when the change degree of the inclination angles of the three crown tips is similar, the +.>Is larger, when there is a large difference in variation between the three crown tips,the value of (2) is small. In the embodiment of the invention, the values of u and v are limited to be [1,3 ]]In other embodimentsThe value range can be adjusted by an implementer according to actual conditions.
The logic of the calculation formula of the probability that the tips and the two-side tips belong to the same broiler chicken is that because the gray values of pixel points on different tips on the same broiler chicken comb are relatively close, the angles between different tips on the same chicken comb are monotonous changes, and the angle difference between the tips on the same chicken comb and the tips on the two sides of the same chicken comb is relatively close; the angle between the first crown tip and the second crown tip is relatively close, but the angle between the second crown tip and the third crown tip is relatively changed. So in combination, the greater the probability that the crown apex and the crown apes at both sides belong to the same broiler chicken, the more likely the d crown apex is to be the crown apex on the same chicken crown as the crown apes at both sides; the smaller the probability that the crown apex and the crown apes at both sides belong to the same broiler chicken, the smaller the probability that the d-th crown apex is the crown apex on the same comb as the crown apes at both sides of the d-th crown apex.
It should be noted that, in the embodiment of the present invention, it may be set that the crown tips are ordered according to the positions of the crown tips in the image, where the image is uniformly divided into n line areas, each line area includes a plurality of lines, and the crown tips are ordered according to the left-to-right order of the abscissa of the vertices of the crown tips in one line area, where if two crown tips corresponding to one abscissa appear, two crown tips of the same abscissa are ordered according to the top-to-bottom order of the ordinate, and in other embodiments, the operator may also order the crown tips according to the actual situation. In the embodiment of the invention, the value of n must be a factor of the number of lines of the image, if the number of lines of the image is not a factor, the number of lines of the image is reduced from small to large until the reduced number of lines of the image is a factor. It should be noted that when one crown peak corresponds to a plurality of probabilities, the corresponding probability is greater than a preset probability threshold, and three crown peaks with the highest probability are divided into the same broiler chicken crown. For example, when the probability that the d-th crown apex and the d-1 th crown apex and the d+1 th crown apex belong to the same broiler chicken is 0.4 and the probability that the d-1 th crown apex and the d+1 th crown apex and the d-2 nd crown apex belong to the same broiler chicken is 0.35, the three crown apes with the largest corresponding probability are divided into the same broiler chicken crowns, and the d-1 th crown apex and the d+1 th crown apex still belong to the same broiler chicken.
So the probability that the crown tips and the crown tips at the two sides belong to the same broiler chicken is obtained by analyzing the crown tips at the d-th crown tip and the crown tips at the two sides; when the probability that the crown tips and the crown tips at two sides belong to the same broiler chicken is larger than or equal to a preset probability threshold value, judging that the crown tips and the crown tips at two sides corresponding to the crown tips belong to the same broiler chicken. In the embodiment of the invention, the preset probability threshold is 0.3, and in other embodiments, the value is adjusted by an implementer according to the actual situation. When the probability that the crown tips and the crown tips at two sides belong to the same broiler chicken is larger than or equal to a preset probability threshold value, reflecting that the probability that the dth crown tip and the crown tips at two sides belong to the same chicken crown is larger; when the probability that the crown tips and the crown tips at two sides belong to the same broiler chicken is smaller than a preset probability threshold, the probability that the d crown tip possibly belongs to the same chicken crown with the crown tips at two sides is reflected to be smaller.
And (3) carrying out the operation on all the cocks in the target cockscomb area, determining the number of cockscombs in the target cockscomb area, and dividing the target cockscomb area according to the cockscombs belonging to the same broiler chicken to obtain a plurality of single cockscomb areas only containing one cockscomb, namely obtaining the cockscombs of each broiler chicken in the target broiler chicken area.
Step S400, detecting the crown tips of each broiler chicken in the target comb region to obtain the health state of the broiler chicken, and carrying out early warning according to the health state of the broiler chicken.
And detecting the crown tips of each broiler chicken in the target comb region to obtain the channel values of the crown tips of each broiler chicken in the RGB three channels in each target comb region.
As the color of the cockscomb is usually bright red, the chickens to be detected are healthier, and when the cockscomb is dark red, infectious diseases such as avian influenza and the like are seen; the cockscomb is purple black and is mostly poisoned by pesticides; cockscomb is bluish purple and may suffer from some toxic diseases; can be generally called as when the deviation between the color of the cockscomb and the bright red is large, the abnormal health of the broiler chickens is reflected.
When the difference between the channel values of the crown tips of the broiler chickens in the target crown region in the RGB three channels and the preset normal crown channel value is larger than a preset difference threshold value, judging that the health state of the broiler chickens is abnormal; and when the difference between the channel values of the crown tips of the broiler chickens in the target crown area in the RGB three channels and the preset normal crown channel value is smaller than or equal to a preset difference threshold value, judging that the health state of the broiler chickens is normal. In the embodiment of the invention, the preset normal cockscomb channel value is the average value of cockscombs of normal chickens in the image set in each channel of RGB, and in other embodiments, the preset normal cockscomb channel value can be obtained through big data statistics by limiting the acquisition method according to actual conditions by an implementer. In the embodiment of the present invention, the preset difference threshold is 20, and in other embodiments, the threshold is adjusted by an implementer according to the actual requirement.
After the health state of the broiler chicken is obtained, early warning is carried out according to the health state of the broiler chicken, and the method specifically comprises the following steps: when the health state of the broiler chickens is abnormal, an alarm signal is sent to the farmers in the farm for early warning.
In summary, the present invention relates to the technical field of image analysis. Firstly, acquiring a monitoring image of broiler chickens; dividing a pending cockscomb area of the broiler monitoring image; determining trend consistency of the pixel points at the edge of each pending cockscomb according to the direction trend situation of the pixel points at the edge of the pending cockscomb in the pending cockscomb area; determining trend pixel points from the pixel points at the edge of the cockscomb to be determined according to the trend consistency; connecting trend pixel points to obtain trend segments, and determining a target cockscomb area based on the number of the trend segments in the cockscomb area to be determined; a line segment connecting the target cockscomb area, and determining a cocktail; determining the crown tips belonging to the same broiler chicken according to the peripheral gray level distribution of the crown tips in the target crown region and the inclination angle change of the crown tips; and detecting the crown tips of each broiler chicken in the target comb region to obtain the health state of the broiler chicken. According to the invention, under the condition that the raising density of the broiler chickens in the raising farm is high, the cockscomb belonging to the same broiler chickens is determined by analyzing the cockscomb, so that the accurate detection of the health state of the broiler chickens is realized.
The embodiment of the invention also provides an artificial intelligence-based broiler raising ecological early warning system, which comprises a memory, a processor and a computer program which is stored in the memory and can run on the processor, wherein the steps of the method are realized when the processor executes the computer program. Because the detailed description is given above for the ecological early warning method for broiler chicken breeding based on artificial intelligence, the detailed description is omitted.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.

Claims (4)

1. The ecological early warning method for broiler chicken breeding based on artificial intelligence is characterized by comprising the following steps of:
acquiring a monitoring image of broiler chickens; dividing a pending cockscomb area of the broiler monitoring image;
determining trend consistency of the pixel points at the edge of each pending cockscomb according to the direction trend situation of the pixel points at the edge of the pending cockscomb in the pending cockscomb area; determining trend pixel points from the pixel points at the edge of the cockscomb to be determined according to the trend consistency; connecting trend pixel points to obtain trend segments, and determining a target cockscomb area based on the number of the trend segments in the cockscomb area to be determined;
a line segment connecting the target cockscomb area, and determining a cocktail; determining the crown tips belonging to the same broiler chicken according to the peripheral gray level distribution of the crown tips in the target crown region and the inclination angle change of the crown tips;
detecting the crown tips of each broiler chicken in the target comb region to obtain the health state of the broiler chicken, and carrying out early warning according to the health state of the broiler chicken;
the method for determining the trend pixel points from the pixel points at the edge of the cockscomb according to the trend consistency comprises the following steps: taking undetermined cockscomb edge pixel points with trend consistency larger than or equal to a preset trend threshold value as trend pixel points;
wherein determining the target cockscomb area based on the number of trend segments in the cockscomb area comprises: determining the smoothness of the region according to the number of trend segments in the cockscomb region to be determined; when the region smoothness is smaller than a preset smoothness threshold value, taking the pending cockscomb region corresponding to the region smoothness as a target cockscomb region;
the ratio of the area of the cockscomb area to the number of trend segments in the cockscomb area is used as the area smoothness of the cockscomb area;
according to the peripheral gray level distribution of the crown tips in the target cockscomb area and the change of the inclination angles of the crown tips, determining the crown tips belonging to the same broiler chicken comprises the following steps:
calculating the probability that the crown tips and the crown tips at two sides belong to the same broiler chicken according to the peripheral gray distribution of the crown tips in the target crown region and the change of the inclination angle of the crown tips;
when the probability that the crown tips and the crown tips at two sides belong to the same broiler chicken is larger than or equal to a preset probability threshold value, judging that the crown tips and the crown tips at two sides corresponding to the crown tips belong to the same broiler chicken;
according to the peripheral gray level distribution of the crown tips in the target cockscomb area and the change of the inclination angles of the crown tips, the calculating of the probability that the crown tips and the crown tips at two sides belong to the same broiler chicken comprises the following steps:
the calculation formula of the probability that the crown tips and the crown tips at two sides belong to the same broiler chicken is as follows:
the method comprises the steps of carrying out a first treatment on the surface of the Wherein (1)>The probability that the (d) th crown apex, the (d-u) th crown apex and the (d+v) th crown apex belong to the same broiler chicken in the (R) th target crown area; />Is a sighard function; min is a function taking the minimum value; max is a maximum function; />The average gray value of each pixel point in the closed area corresponding to the d-th crown peak is obtained; />The average gray value of each pixel point in the closed area corresponding to the d-u crown points; />The average gray value of each pixel point in the closed area corresponding to the d+v crown points; />Is the inclination angle of the (d+v) th crown tip; />Is the inclination angle of the d-th crown tip; />Is the inclination angle of the d-u crown tip; />Is a sign function;
the determining the trend consistency of the pixel points at the edge of the cockscomb according to the direction trend situation of the pixel points at the edge of the cockscomb in the cockscomb area comprises the following steps:
acquiring gradient direction values corresponding to pixel points at the edge of each pending cockscomb in the pending cockscomb area;
taking the left-most cockscomb edge pixel point in each cockscomb area as a chain code starting point corresponding to each cockscomb area, and constructing an edge direction chain code corresponding to the cockscomb area by using a gradient direction value corresponding to each cockscomb edge pixel point;
and taking any element in the edge direction chain code as a target element, and taking the number of elements which are the same as the numerical value of the target element in a sliding window corresponding to the target element as the trend consistency of the pixel points of the edge of the cockscomb to be determined corresponding to the target element.
2. The artificial intelligence based broiler raising ecological early warning method of claim 1, wherein the line segment connecting the target cockscomb area, determining the cocktail, comprises:
connecting line segments in the target cockscomb area to obtain at least two undetermined angles;
and when the undetermined angle is an acute angle and the inner area of the undetermined angle belongs to the target cockscomb area, taking the undetermined angle as a cockscomb tip.
3. The artificial intelligence-based ecological early warning method for broiler chicken raising of claim 1, wherein the dividing the pending cockscomb area of the broiler chicken monitoring image comprises:
and carrying out threshold segmentation on the broiler monitoring image, and marking the region obtained by threshold segmentation as a cocked comb region.
4. An artificial intelligence-based broiler raising ecological early warning system, comprising a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the steps of the artificial intelligence-based broiler raising ecological early warning method according to any one of claims 1-3 are realized when the processor executes the computer program.
CN202311743590.2A 2023-12-19 2023-12-19 Ecological early warning method and system for broiler chicken breeding based on artificial intelligence Active CN117422716B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311743590.2A CN117422716B (en) 2023-12-19 2023-12-19 Ecological early warning method and system for broiler chicken breeding based on artificial intelligence

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311743590.2A CN117422716B (en) 2023-12-19 2023-12-19 Ecological early warning method and system for broiler chicken breeding based on artificial intelligence

Publications (2)

Publication Number Publication Date
CN117422716A CN117422716A (en) 2024-01-19
CN117422716B true CN117422716B (en) 2024-03-08

Family

ID=89532864

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311743590.2A Active CN117422716B (en) 2023-12-19 2023-12-19 Ecological early warning method and system for broiler chicken breeding based on artificial intelligence

Country Status (1)

Country Link
CN (1) CN117422716B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118155142B (en) * 2024-05-09 2024-08-16 浙江大华技术股份有限公司 Object density recognition method and event recognition method
CN119205767B (en) * 2024-11-27 2025-03-04 陕西威达机械制造股份有限公司 CNC system-based intelligent positioning method for workpiece drilling

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113256703A (en) * 2021-06-03 2021-08-13 河北农业大学 Method and system for determining shape of cockscomb based on image processing
WO2022089236A1 (en) * 2020-11-02 2022-05-05 腾讯科技（深圳）有限公司 Image processing method apparatus based on artificial intelligence, and computer device and storage medium
CN114998321A (en) * 2022-07-19 2022-09-02 南通博纳纺织品有限公司 Textile material surface hairiness degree identification method based on optical means
KR20220129808A (en) * 2021-03-17 2022-09-26 박만기 Nets for aquaculture of cockroaches
CN115861721A (en) * 2023-02-28 2023-03-28 山东大佳机械有限公司 Livestock and poultry breeding spraying equipment state identification method based on image data
CN116912260A (en) * 2023-09-15 2023-10-20 沂水友邦养殖服务有限公司 Broiler chicken breeding health state detection method based on artificial intelligence

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7146057B2 (en) * 2002-07-10 2006-12-05 Northrop Grumman Corporation System and method for image analysis using a chaincode
CN108921823B (en) * 2018-06-08 2020-12-01 Oppo广东移动通信有限公司 Image processing method, apparatus, computer-readable storage medium and electronic device

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2022089236A1 (en) * 2020-11-02 2022-05-05 腾讯科技（深圳）有限公司 Image processing method apparatus based on artificial intelligence, and computer device and storage medium
KR20220129808A (en) * 2021-03-17 2022-09-26 박만기 Nets for aquaculture of cockroaches
CN113256703A (en) * 2021-06-03 2021-08-13 河北农业大学 Method and system for determining shape of cockscomb based on image processing
CN114998321A (en) * 2022-07-19 2022-09-02 南通博纳纺织品有限公司 Textile material surface hairiness degree identification method based on optical means
CN115861721A (en) * 2023-02-28 2023-03-28 山东大佳机械有限公司 Livestock and poultry breeding spraying equipment state identification method based on image data
CN116912260A (en) * 2023-09-15 2023-10-20 沂水友邦养殖服务有限公司 Broiler chicken breeding health state detection method based on artificial intelligence

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Development of an early warning algorithm to detect sick broilers;Xiaolin Zhuang 等;《Computers and Electronics in Agriculture》;20180131;第第144卷卷;第102-113页 *
基于支持向量机的养鸡场死鸡检测方法研究;彭彦松 等;《知网》;20100815;第2010年卷(第8期);全文 *
基于鸡头特征的病鸡识别方法研究;毕敏娜;张铁民;庄晓霖;焦培荣;;农业机械学报;20180125(01);全文 *

Also Published As

Publication number Publication date
CN117422716A (en) 2024-01-19

Similar Documents

Publication Publication Date Title
CN117422716B (en) Ecological early warning method and system for broiler chicken breeding based on artificial intelligence
EP3553742B1 (en) Method and device for identifying pathological picture
Gregson et al. Automated grading of venous beading
US20190014982A1 (en) Automated blood vessel feature detection and quantification for retinal image grading and disease screening
Abbasi-Sureshjani et al. Automatic detection of vascular bifurcations and crossings in retinal images using orientation scores
Omar et al. Detection and classification of retinal fundus images exudates using region based multiscale LBP texture approach
US9672610B2 (en) Image processing apparatus, image processing method, and computer-readable recording medium
Barbosa et al. Towards automatic quantification of the epicardial fat in non-contrasted CT images
US12293576B2 (en) Determining type of to-be-classified image based on signal waveform graph
CN116228764B (en) Neonate disease screening blood sheet acquisition quality detection method and system
CN117974692B (en) Ophthalmic medical image processing method based on region growing
CN116912260B (en) Broiler chicken breeding health state detection method based on artificial intelligence
CN104573716A (en) Eye fundus image arteriovenous retinal blood vessel classification method based on breadth first-search algorithm
CN118247277B (en) Self-adaptive enhancement method for lung CT image
Bharali et al. Detection of hemorrhages in diabetic retinopathy analysis using color fundus images
Zhou et al. Automatic optic disc boundary extraction based on saliency object detection and modified local intensity clustering model in retinal images
CN119600086A (en) Intelligent detection method and system for cross-sectional ovality of medical wires
CN117876801B (en) Method for predicting diabetic nephropathy based on fundus blood vessel characteristics and artificial intelligence
KR102220574B1 (en) Method, apparatus and computer program for calculating quality score threshold for filtering image data
Soares et al. Exudates dynamic detection in retinal fundus images based on the noise map distribution
CN114627194B (en) LED lamp luminous state detection method and device
Park et al. Diabetic retinopathy classification using c4. 5
CN118537896B (en) Livestock and veterinary animal body temperature intelligent detection system
Roy Chowdhury et al. Towards an automated approach to the detection of retinal abnormalities
CN113421218B (en) Extraction method of vascular network branch point

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant