[go: up one dir, main page]

CN112215104A - A method, device and device for sea ice extraction based on superpixel segmentation - Google Patents

A method, device and device for sea ice extraction based on superpixel segmentation Download PDF

Info

Publication number
CN112215104A
CN112215104A CN202011042927.3A CN202011042927A CN112215104A CN 112215104 A CN112215104 A CN 112215104A CN 202011042927 A CN202011042927 A CN 202011042927A CN 112215104 A CN112215104 A CN 112215104A
Authority
CN
China
Prior art keywords
superpixel
pixel
sea ice
area
entropy
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.)
Pending
Application number
CN202011042927.3A
Other languages
Chinese (zh)
Inventor
王芳
郑慧
关峰
曾杰
孙晓月
郭榕刚
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.)
Guojiao Space Information Technology Beijing Co ltd
Original Assignee
Guojiao Space Information Technology Beijing 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 Guojiao Space Information Technology Beijing Co ltd filed Critical Guojiao Space Information Technology Beijing Co ltd
Priority to CN202011042927.3A priority Critical patent/CN112215104A/en
Publication of CN112215104A publication Critical patent/CN112215104A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes
    • G06V20/13—Satellite images
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; 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 OR CALCULATING; 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 OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/40—Analysis of texture
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/90—Determination of colour characteristics
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30181—Earth observation

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Multimedia (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Astronomy & Astrophysics (AREA)
  • Remote Sensing (AREA)
  • Image Analysis (AREA)
  • Image Processing (AREA)

Abstract

本发明的实施例提供了一种基于超像素分割的海冰提取方法、装置和设备。所述方法包括对原始遥感图像的非陆地部分，采用SLIC的超像素分割算法进行超像素分割；根据超像素分割结果，基于灰度共生矩阵，计算每个超像素的对比度和熵的均值，生成基于超像素的特征图像；采用Otsu阈值检测方法计算对比度阈值和熵阈值，对海水区域和海冰区域进行分类，并提取所述海冰区域。以此方式，能够实现海冰的自动化提取，为海上航运和海上石油开采等行业的海上活动提取参考依据，而且海冰作为全球气候变化的重要参与者，海冰提取进一步为气候变化研究提供了技术支持。

Figure 202011042927

Embodiments of the present invention provide a method, apparatus and device for sea ice extraction based on superpixel segmentation. The method includes using the superpixel segmentation algorithm of SLIC to perform superpixel segmentation on the non-land part of the original remote sensing image; according to the superpixel segmentation result, based on the gray level co-occurrence matrix, calculate the average value of the contrast and entropy of each superpixel, and generate Based on the feature image of superpixels; using the Otsu threshold detection method to calculate the contrast threshold and the entropy threshold, classify the seawater area and the sea ice area, and extract the sea ice area. In this way, the automatic extraction of sea ice can be realized, and the reference basis for marine activities in industries such as marine shipping and offshore oil extraction can be extracted. Moreover, as an important participant in global climate change, sea ice extraction further provides climate change research. Technical Support.

Figure 202011042927

Description

Sea ice extraction method, device and equipment based on superpixel segmentation
Technical Field
The embodiment of the invention relates to the application field of remote sensing images in general, and more particularly relates to a sea ice extraction method, device and equipment based on superpixel segmentation.
Background
The icing sea area in China is mainly located in the Bohai sea and is mostly concentrated in the Liaodong bay in the north of the icing sea area. In winter each year, the development of marine resources, the design of port and coast engineering, the transportation of marine vehicles and other human activities on the sea are affected by the formation, movement and change of sea ice to different degrees, and particularly, serious ice conditions have great harm to offshore oil platforms and other offshore structures. With the development of remote sensing technology, the monitoring of sea ice by using multispectral images and radar images becomes an important research content. In the sea ice extraction based on the remote sensing image, the currently more common method mainly comprises the steps of adopting the sea ice extraction based on the brightness temperature and the reflectivity characteristics of the MODIS image, and searching the boundary value of ice and water by counting the histogram of the brightness temperature and the reflectivity characteristics of the multispectral image to extract the sea ice; adopting a maximum likelihood sea ice extraction method based on gray scale and textural features of an SAR image; and (3) adopting a sea ice extraction method based on texture features and an SVM classification method of the SAR image, and the like.
By adopting the sea ice extraction method based on the brightness temperature and the reflectivity characteristics of the MODIS image, the influence of interferents such as clouds and the like is difficult to completely remove, the sea ice and the sea water area are difficult to distinguish in partial areas, the sea ice extraction work belongs to semi-automatic extraction, and more manual intervention is performed. The maximum likelihood sea ice extraction method based on the gray scale and the texture features of the SAR image needs a large amount of training sample data and is serious in extraction confusion of broken ice and open water with approximate gray scale values. The sea ice extraction method based on the texture features and the SVM classification method of the SAR image also needs a large amount of training sample data, and more identification errors exist in the area with large wind speed change.
Disclosure of Invention
According to an embodiment of the present invention, a sea ice extraction scheme based on superpixel segmentation is provided.
In a first aspect of the invention, a method for extracting sea ice based on superpixel segmentation is provided. The method comprises the following steps: obtaining a non-land part of an original remote sensing image;
performing superpixel segmentation on the non-land part of the original remote sensing image by adopting a superpixel segmentation algorithm of SLIC (narrow-line segmentation algorithm);
calculating the mean value of the contrast and entropy of each super pixel based on the gray level co-occurrence matrix according to the super pixel segmentation result, and generating a characteristic image based on the super pixels;
and calculating a contrast threshold and an entropy threshold by adopting an Otsu threshold detection method, classifying the seawater area and the sea ice area, and extracting the sea ice area.
Further, for the non-land part of the original remote sensing image, performing superpixel segmentation by using a superpixel segmentation algorithm of the SLIC, comprising:
step 1: setting the side length of the super pixel to generate a plurality of initial super pixel seed points;
step 2: setting a corresponding adjacent area by taking an initial super-pixel seed point as a center, calculating the gradient value of each point in the adjacent area, and taking the point with the minimum gradient value as the super-pixel seed point of the adjacent area;
and step 3: setting a pixel range, and calculating the characteristic distance from each pixel point in the pixel range to the current super-pixel seed point by taking each super-pixel seed point as a center; if the characteristic distance from the current pixel point to the current super-pixel seed point is smaller than the characteristic distance from the current pixel point to the super-pixel seed point, the relationship of the current pixel point is adjusted to the current super-pixel seed point;
and 4, step 4: calculating the center of the super pixel of each adjusted super pixel seed point to serve as the seed point of the corresponding super pixel;
and 5: and (5) iterating the step (3) and the step (4) until the preset iteration times are reached.
Further, the characteristic distance from each pixel point to the current super-pixel seed point is as follows:
Figure BDA0002707156250000031
wherein N isiThe gray value of the current super pixel seed point is obtained; n is a radical ofjThe gray value of the current pixel point is obtained; xiThe abscissa of the current superpixel seed point; y isiIs the ordinate of the current superpixel seed point; xjThe abscissa of the current pixel point is; y isjIs the ordinate of the current pixel point.
Further, the calculating the mean of the contrast and the entropy of each super pixel and generating the super pixel-based feature image comprises:
compressing the gray level of the original remote sensing image, and setting a first sliding window to calculate a gray level co-occurrence matrix;
calculating the contrast and entropy of the pixels according to the horizontal gray level co-occurrence matrix of each pixel;
and calculating the average value of the contrast and the entropy of a plurality of pixel points in each super pixel, taking the average value as the contrast characteristic value and the entropy characteristic value of the current super pixel, and generating the characteristic image based on the super pixel.
Further, the contrast and entropy of the pixel are:
Figure BDA0002707156250000032
Figure BDA0002707156250000033
wherein CON is the contrast of the pixel; ENT is the entropy of the pixel; k is the size of the first sliding window; g (i, j) is the value of the ith row and j column of the gray level co-occurrence matrix.
Further, the classifying the sea water area and the sea ice area includes:
if the contrast value of the super pixel is greater than the contrast threshold and/or the entropy value of the super pixel is greater than the entropy threshold, the region where the super pixel is located is a sea ice region; otherwise, it is a seawater area.
Further, still include:
setting a second sliding window, and calculating the sea ice proportion of the seawater area and the sea ice area in the second sliding window to generate a sea ice density characteristic image;
setting a sea ice density of a maximum connected region of the sea water regions to 0;
and identifying the area with the sea ice density larger than 0 in the sea ice density characteristic image as the sea ice, and taking the sea ice identification result as a sea ice extraction result.
In a second aspect of the invention, a sea ice extraction device based on superpixel segmentation is provided. The device includes:
the separation module is used for separating the land part and the non-land part in the original remote sensing image to obtain the non-land part of the original remote sensing image;
the super-pixel segmentation module is used for carrying out super-pixel segmentation on the non-land part of the original remote sensing image by adopting a super-pixel segmentation algorithm of SLIC (narrow line capacitance);
the characteristic image generation module is used for calculating the mean value of the contrast and the entropy of each super pixel based on the gray level co-occurrence matrix according to the super pixel segmentation result and generating a characteristic image based on the super pixels;
and the classification extraction module is used for calculating a contrast threshold and an entropy threshold of the characteristic image by adopting an Otsu threshold detection method, classifying the sea water area and the sea ice area and extracting the sea ice area.
In a third aspect of the invention, an electronic device is provided. The electronic device includes: a memory having a computer program stored thereon and a processor implementing the method as described above when executing the program.
In a fourth aspect of the invention, a computer-readable storage medium is provided, on which a computer program is stored which, when being executed by a processor, carries out the method as according to the first aspect of the invention.
It should be understood that the statements herein reciting aspects are not intended to limit the critical or essential features of any embodiment of the invention, nor are they intended to limit the scope of the invention. Other features of the present invention will become apparent from the following description.
According to the invention, the sea ice is extracted by adopting a super-pixel segmentation based method, so that the sea ice extraction precision is improved; the texture features of the gray level co-occurrence matrix are used as classification features, and the texture mean value of the super-pixel region is adopted for classifying the seawater and the sea ice, so that the condition that classification results are scattered due to pixel-based classification is avoided, and the integrity of the classification results is effectively ensured; and the sea water and the sea ice are automatically classified by adopting a self-adaptive threshold value method, so that the automatic classification can be realized for different SAR images, the manual intervention is not needed, the automatic extraction of the sea ice is realized, and the automatic level of the sea ice extraction is improved.
Drawings
The above and other features, advantages and aspects of various embodiments of the present invention will become more apparent by referring to the following detailed description when taken in conjunction with the accompanying drawings. In the drawings, like or similar reference characters designate like or similar elements, and wherein:
FIG. 1 shows a flow chart of a method of sea ice extraction based on superpixel segmentation in accordance with the present invention;
FIG. 2 shows a schematic diagram of a non-terrestrial portion of an original remote-sensing image according to the invention;
FIG. 3 illustrates a superpixel segmentation flow diagram according to an embodiment of the present invention;
FIG. 4 shows a diagram of a superpixel segmentation result according to an embodiment of the present invention;
FIG. 5 illustrates a schematic of a neighborhood of super-pixel seed points, according to an embodiment of the present invention;
FIG. 6 shows a schematic diagram of a contrast feature image according to an embodiment of the invention;
FIG. 7 shows a schematic diagram of an entropy feature image according to an embodiment of the invention;
FIG. 8 shows a block diagram of a super-pixel segmentation based sea ice extraction apparatus according to an embodiment of the present invention;
FIG. 9 illustrates a block diagram of an exemplary electronic device capable of implementing embodiments of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
In addition, the term "and/or" herein is only one kind of association relationship describing an associated object, and means that there may be three kinds of relationships, for example, a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone. In addition, the character "/" herein generally indicates that the former and latter related objects are in an "or" relationship.
According to the invention, the sea ice is extracted by adopting a super-pixel segmentation based method, so that the sea ice extraction precision is improved; the texture features of the gray level co-occurrence matrix are used as classification features, and the texture mean value of the super-pixel region is adopted for classifying the seawater and the sea ice, so that the condition that classification results are scattered due to pixel-based classification is avoided, and the integrity of the classification results is effectively ensured; and the sea water and the sea ice are automatically classified by adopting a self-adaptive threshold value method, so that the automatic classification can be realized for different SAR images, the manual intervention is not needed, the automatic extraction of the sea ice is realized, and the automatic level of the sea ice extraction is improved.
Fig. 1 shows a flow chart of a method for sea ice extraction based on superpixel segmentation according to the present invention.
Firstly, separating a land part and a non-land part in an original remote sensing image to obtain the non-land part of the original remote sensing image.
The original remote sensing image comprises a land part and a non-land part, and the original remote sensing image needs to be processed firstly, namely the land part is separated from the non-land part; and (3) cutting off a land part from the original remote sensing data by using the known coastline vector data, and leaving a non-land part in the cut original remote sensing data as shown in figure 2.
The sea ice extraction method based on the superpixel segmentation comprises the following steps:
s100, performing superpixel segmentation on the non-land part of the original remote sensing image by using a superpixel segmentation algorithm of a Simple Linear Iterative Clustering (SLIC).
The process of super-pixel segmentation using SLIC is shown in fig. 3, and includes:
s102, setting the side length of the super pixel to generate a plurality of initial super pixel seed points.
Firstly, the size S of the super-pixel needs to be set, i.e. the side length of the super-pixel in a regular case, the size of S can be selected empirically or set by input, for example, S is 10.
Secondly, a number of initial superpixel seed points, e.g. Z, are generated uniformly according to the size S of the selected superpixel1,Z2,…,Zi,…,Zk。
As an embodiment of the present invention, the side length of the super pixel is set to 10, and the coordinates of the generated initial super pixel seed points are (5,5), (5,15), (5,25) … … (15,5), (15,15), (15,25) … ….
S104, setting a corresponding adjacent area by taking an initial super-pixel seed point as a center, calculating the gradient value of each point in the adjacent area, and taking the point with the minimum gradient value as the super-pixel seed point of the adjacent area.
To avoid the super-pixel seed point falling on noise or a boundary, the position of the initial super-pixel seed point needs to be adjusted.
Setting a proximity region for each initial superpixel seed point, wherein the proximity region takes the corresponding initial superpixel seed point as a center; as shown in fig. 5, the shaded area is an area of a 3 × 3 range centered on the initial super-pixel seed point (5,5), that is, (4,4), (4,5), (4,6), (5,4), (5,5), (5,6), (6,4), (6,5), (6, 6).
Calculating the gradient value of each point in the adjacent area, and taking the point with the minimum gradient value as a super-pixel seed point of the adjacent area;
in the embodiment of the present invention, the gradient values of the respective points (4,4), (4,5), (4,6), (5,4), (5,5), (5,6), (6,4), (6,5) and (6,6) are calculated, for example, the gradient value of (4,6) is the smallest, and then (4,6) is adjusted and moved from (5,5) to (4,6) as the super-pixel seed point of the neighboring region, i.e., the super-pixel seed point, so as to avoid the situation that the super-pixel seed point falls on a noise point or a boundary.
S106, setting a pixel range, and calculating the characteristic distance from each pixel point in the pixel range to the current super-pixel seed point by taking each super-pixel seed point as a center.
The pixel range is set to (2S +1) × (2S +1), for example, when the side length of the super pixel is 10, the pixel range is the super pixel seed point ZiA central 21 × 21 pixel range.
Respectively calculating each pixel point in the pixel range to the current super-pixel seed point ZiThe characteristic distance of (c). In calculating the feature distance, three dimensions of the gray levels N and XY are considered.
The characteristic distance calculation formula is as follows:
Figure BDA0002707156250000081
wherein D isijIs a point P in the pixel rangejTo the current superpixel seed point ZiThe characteristic distance of (d); n is a radical ofiThe gray value of the current super pixel seed point is obtained; n is a radical ofjThe gray value of the current pixel point is obtained; xiThe abscissa of the current superpixel seed point; y isiIs the ordinate of the current superpixel seed point; xjThe abscissa of the current pixel point is; y isjIs the ordinate of the current pixel point.
If the characteristic distance from the current pixel point to the current super-pixel seed point is smaller than the characteristic distance from the current pixel point to the super-pixel seed point, the relationship of the current pixel point is adjusted to the current super-pixel seed point; otherwise, no adjustment is carried out, namely the current pixel point is thrown to the original super-pixel seed point.
As an embodiment of the invention, the pixel point (6,5) belongs to a super-pixel seed point (4,6), but the characteristic distance D from the pixel point (6,5) to another super-pixel seed point (7,6)1Less than its characteristic distance D to (4,6)2Then pixel (6,5) is adjusted to super pixel (7, 6).
And S108, calculating the center of the super pixel of each adjusted super pixel seed point to serve as the seed point of the corresponding super pixel.
Since the pixel point belonging to each super-pixel seed point is adjusted, so that the super-pixel seed point is not the central point of the pixel point belonging to the super-pixel seed point, the average value of the abscissa and the average value of the ordinate of the pixel point belonging to the current super-pixel seed point need to be calculated, and the coordinate of the central point is obtained and used as the seed point of the corresponding super-pixel.
And S110, iterating S206 and S208 until the preset iteration number is reached.
S106 and S108 complete a round of adjustment of the super-pixel seed points and the pixels belonging to the super-pixel seed points, and the super-pixel seed points and the pixels belonging to the super-pixel seed points can be adjusted into a more accurate super-pixel segmentation result by a plurality of times of adjustment. Therefore, the iteration times are preset, and generally set to 10, that is, after 10 rounds of S106 and S108 adjustments are completed, the superpixel segmentation result can be obtained.
S200, calculating the mean value of the contrast and the entropy of each super pixel based on the gray level co-occurrence matrix according to the super pixel segmentation result, and generating the characteristic image based on the super pixels.
According to the method, the texture features based on the gray level co-occurrence matrix are adopted to extract the texture features of the super pixels.
In the texture feature calculation process based on the gray level co-occurrence matrix, the influence of the gray level of the image on the calculation speed is very large, and in order to reduce the calculation amount, the gray level of the image needs to be compressed.
The S200 includes the steps of:
s202, compressing the gray level of the original remote sensing image, and setting a first sliding window to calculate a gray level co-occurrence matrix.
As an embodiment of the invention, the original remote sensing image has a gray level of 256, and the gray level can be compressed to 32 levels. The specific gray level compression process is as follows:
firstly, image histogram equalization is carried out, then the gray scale is divided by 8, and after the integration, the 0-255 gray scale can be converted into the 0-31 gray scale.
The first sliding window can adopt a sliding window with the size of 11 multiplied by 11 to calculate a horizontal gray level co-occurrence matrix; experiments prove that the horizontal gray level co-occurrence matrix calculated by adopting the 11 multiplied by 11 sliding window has better effect on extracting the sea ice.
S204, calculating the contrast and entropy of the pixels according to the horizontal gray level co-occurrence matrix of each pixel:
the contrast and entropy of the pixel are:
Figure BDA0002707156250000101
Figure BDA0002707156250000102
wherein CON is the contrast of the pixel; ENT is the entropy of the pixel; k is the size of the first sliding window; g (i, j) is the value of the ith row and j column of the gray level co-occurrence matrix.
S206, calculating the average value of the contrast CON and the entropy ENT of a plurality of pixel points in each super pixel, taking the average value as the contrast characteristic value and the entropy characteristic value of the current super pixel, and generating the characteristic image based on the super pixel.
In the process of S204, the contrast and entropy of each pixel are obtained, and mean value calculation of the contrast and entropy is performed on the pixels belonging to the same superpixel to obtain a contrast characteristic value and an entropy characteristic value of the corresponding superpixel, and generate a contrast characteristic image, as shown in fig. 6; an entropy feature image is generated as shown in fig. 7.
The sea water and sea ice are classified by adopting the texture mean value of the super-pixel area, so that the scattering of classification results caused by pixel-based classification is avoided, and the integrity of the classification results is effectively ensured.
S300, calculating a contrast threshold and an entropy threshold of the characteristic image by adopting an Otsu threshold detection method, classifying a seawater area and a sea ice area, and extracting the sea ice area.
The method 300 comprises the following steps:
s302, determining optimal thresholds of the contrast and the entropy of the two texture features by adopting an Otsu threshold detection method to obtain a contrast threshold T1 and an entropy threshold T2.
And the optimal threshold is determined by adopting an Otsu threshold detection method, so that automatic classification can be realized for different SAR images without manual intervention.
S304, classifying the seawater area and the sea ice area, and extracting the sea ice area.
According to experience, the contrast value of the sea ice is larger than that of the sea water, and the entropy value of the sea ice is larger than that of the sea water. And judging each super pixel by adopting thresholds T1 and T2 of two characteristics of contrast and entropy, and classifying the sea surface into seawater or sea ice, namely finishing the extraction of the sea ice.
The specific judgment process comprises the following steps: if the contrast value of the super pixel is greater than the contrast threshold and/or the entropy value of the super pixel is greater than the entropy threshold, the region where the super pixel is located is a sea ice region; otherwise, it is a seawater area.
Therefore, if the contrast value of the superpixel is greater than the contrast threshold and the entropy value of the superpixel is greater than the entropy threshold, the superpixel is a sea ice region;
if the contrast value of the super pixel is greater than the contrast threshold value and the entropy value of the super pixel is less than the entropy threshold value, the super pixel is a sea ice area;
if the contrast value of the super pixel is smaller than the contrast threshold value and the entropy value of the super pixel is larger than the entropy threshold value, the super pixel is a sea ice area;
and if the contrast value of the super pixel is smaller than the contrast threshold value and the entropy value of the super pixel is smaller than the entropy threshold value, determining that the super pixel is a seawater area.
In conclusion, the classification of the seawater area and the sea ice area is realized, and the sea ice area can be extracted.
As a preferred embodiment of the present invention, the sea ice region extracted in step S300 may be optimized by calculating the sea ice density.
The optimization process S400 includes:
s402, setting a second sliding window, calculating the sea ice proportion of the seawater area and the sea ice area in the second sliding window, and generating a sea ice density characteristic image.
The second sliding window may be set to an 81 × 81 sliding window.
S404, setting the sea ice density of the largest communication area in the sea water area to be 0.
The maximum connected region can be identified by calculating the number of pixel points contained in the seawater region, namely if the number of pixel points contained in one connected region is more than the number of pixel points contained in other connected regions, the connected region is the maximum connected region, and the sea ice density of the region is set to be 0, namely the region is considered as the seawater region.
S406, identifying the area with the sea ice density larger than 0 in the sea ice density characteristic image as the sea ice, and taking the sea ice identification result as a sea ice extraction result.
The area with the sea ice density equal to 0 is identified as sea water, the area with the sea ice density greater than 0 is identified as sea ice, the optimization of sea ice extraction is completed, the extraction precision of crushed ice such as primary ice is improved, and the sea ice edge precision is high.
According to the embodiment of the invention, the sea ice is extracted by adopting a method based on superpixel segmentation, so that the precision of sea ice extraction is improved; the texture features of the gray level co-occurrence matrix are used as classification features, and the texture mean value of the super-pixel region is adopted for classifying the seawater and the sea ice, so that the condition that classification results are scattered due to pixel-based classification is avoided, and the integrity of the classification results is effectively ensured; and the sea water and the sea ice are automatically classified by adopting a self-adaptive threshold value method, so that the automatic classification can be realized for different SAR images, the manual intervention is not needed, the automatic extraction of the sea ice is realized, and the automatic level of the sea ice extraction is improved.
It should be noted that, for simplicity of description, the above-mentioned method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the present invention is not limited by the order of acts, as some steps may occur in other orders or concurrently in accordance with the invention. Further, those skilled in the art should also appreciate that the embodiments described in the specification are exemplary embodiments and that the acts and modules illustrated are not necessarily required to practice the invention.
The above is a description of method embodiments, and the embodiments of the present invention are further described below by way of apparatus embodiments.
As shown in fig. 8, the apparatus 800 includes:
and the separation module 810 is configured to separate a land part and a non-land part in the original remote sensing image to obtain the non-land part of the original remote sensing image.
And a super-pixel segmentation module 820, configured to perform super-pixel segmentation on the non-land portion of the original remote sensing image by using a super-pixel segmentation algorithm of the SLIC.
The super-pixel segmentation module 820 further comprises:
an initial super-pixel seed point generating module 821, configured to set a side length of a super pixel and generate a plurality of initial super-pixel seed points;
a first adjusting module 822, configured to set a corresponding neighboring area with an initial super-pixel seed point as a center, calculate a gradient value of each point in the neighboring area, and use a point with a minimum gradient value as the super-pixel seed point of the neighboring area;
a characteristic distance calculating module 823, configured to set a pixel range, and calculate a characteristic distance from each pixel point in the pixel range to a current super-pixel seed point by taking each super-pixel seed point as a center; the characteristic distance from each pixel point to the current super-pixel seed point is as follows:
Figure BDA0002707156250000131
wherein N isiThe gray value of the current super pixel seed point is obtained; n is a radical ofjThe gray value of the current pixel point is obtained; xiThe abscissa of the current superpixel seed point; y isiIs the ordinate of the current superpixel seed point; xjThe abscissa of the current pixel point is; y isjIs the ordinate of the current pixel point.
A second adjusting module 824, configured to determine that, if the feature distance from the current pixel point to the current super-pixel seed point is smaller than the feature distance from the current pixel point to the corresponding super-pixel seed point, the relationship of the current pixel point is adjusted to the current super-pixel seed point;
a third adjusting module 825, configured to calculate, for each adjusted super-pixel seed point, a super-pixel center of the super-pixel seed point as a seed point corresponding to the super-pixel;
the iteration module 826 is configured to preset iteration times, and perform iteration calling on the second adjustment module 824 and the third adjustment module 825, where after each iteration, the iteration times are increased by 1 until the preset iteration times are reached.
And a feature image generation module 830, configured to calculate, according to the super-pixel segmentation result, a mean value of the contrast and the entropy of each super-pixel based on the gray level co-occurrence matrix, and generate a feature image based on the super-pixels.
The feature image generation module 830 further includes:
the first calculating module 831 is used for compressing the gray level of the original remote sensing image and setting a first sliding window to calculate a gray level co-occurrence matrix;
a second calculating module 832 for calculating the contrast and entropy of the pixels according to the horizontal gray level co-occurrence matrix of each pixel; the contrast and entropy of the pixel are:
Figure BDA0002707156250000141
Figure BDA0002707156250000142
wherein CON is the contrast of the pixel; ENT is the entropy of the pixel; k is the size of the first sliding window; g (i, j) is the value of the ith row and j column of the gray level co-occurrence matrix.
The third calculating module 833 is configured to calculate a mean value of contrast CON and entropy ENT of a plurality of pixel points in each super pixel, where the mean value is used as a contrast characteristic value and an entropy characteristic value of the current super pixel, and generate a characteristic image based on the super pixel.
The classification extraction module 840 is configured to calculate a contrast threshold and an entropy threshold by using an Otsu threshold detection method, classify the sea water region and the sea ice region, and extract the sea ice region.
If any one of the following two conditions is met, the area where the super pixel is located is a sea ice area; otherwise, the area where the super pixel is located is a seawater area; obtaining a sea water area and a sea ice area:
the first condition is: the contrast value of the superpixel is greater than the contrast threshold;
the second condition is: the entropy value of the superpixel is greater than the entropy threshold.
The apparatus 800 also includes an optimization extraction module 850.
The optimized extraction module 850 further includes:
a fourth calculating module 851, configured to set a second sliding window, calculate a sea ice ratio of the sea water area and the sea ice area in the second sliding window, and generate a sea ice density feature image;
a setting module 852 for setting the sea ice density of the largest connected region in the sea water region to 0;
an identifying module 853, configured to identify, as the sea ice, an area in the sea ice density feature image where the sea ice density is greater than 0, and take a sea ice identification result as a sea ice extraction result.
It can be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working process of the described module may refer to the corresponding process in the foregoing method embodiment, and is not described herein again.
As shown in fig. 9, the electronic device includes a Central Processing Unit (CPU) that can perform various appropriate actions and processes according to computer program instructions stored in a Read Only Memory (ROM) or computer program instructions loaded from a storage unit into a Random Access Memory (RAM). In the RAM, various programs and data required for the operation of the device can also be stored. The CPU, ROM, and RAM are connected to each other via a bus. An input/output (I/O) interface is also connected to the bus.
A plurality of components in an electronic device are connected to an I/O interface, including: an input unit such as a keyboard, a mouse, etc.; an output unit such as various types of displays, speakers, and the like; storage units such as magnetic disks, optical disks, and the like; and a communication unit such as a network card, modem, wireless communication transceiver, etc. The communication unit allows the electronic device to exchange information/data with other devices via a computer network such as the internet and/or various telecommunication networks.
The processing unit performs the various methods and processes described above, such as methods S100-S400. For example, in some embodiments, the methods S100-S400 may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and/or installed onto the device via ROM and/or the communication unit. When the computer program is loaded into RAM and executed by the CPU, one or more of the steps of methods S100-S400 described above may be performed. Alternatively, in other embodiments, the CPU may be configured to perform methods S100-S400 by any other suitable means (e.g., by way of firmware).
The functions described herein above may be performed, at least in part, by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: a Field Programmable Gate Array (FPGA), an Application Specific Integrated Circuit (ASIC), an Application Specific Standard Product (ASSP), a system on a chip (SOC), a load programmable logic device (CPLD), and the like.
Program code for implementing the methods of the present invention may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions/operations specified in the flowchart and/or block diagram to be performed. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.
In the context of the present invention, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
Further, while operations are depicted in a particular order, this should be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. Under certain circumstances, multitasking and parallel processing may be advantageous. Likewise, while several specific implementation details are included in the above discussion, these should not be construed as limitations on the scope of the invention. Certain features that are described in the context of separate embodiments can also be implemented in combination in a single implementation. Conversely, various features that are described in the context of a single implementation can also be implemented in multiple implementations separately or in any suitable subcombination.
Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.

Claims (10)

1.一种基于超像素分割的海冰提取方法，其特征在于，包括：1. a sea ice extraction method based on superpixel segmentation, is characterized in that, comprises: 对原始遥感图像的非陆地部分，采用SLIC的超像素分割算法进行超像素分割；For the non-land parts of the original remote sensing image, the superpixel segmentation algorithm of SLIC is used for superpixel segmentation; 根据超像素分割结果，基于灰度共生矩阵，计算每个超像素的对比度和熵的均值，生成基于超像素的特征图像；According to the superpixel segmentation results, based on the gray level co-occurrence matrix, calculate the mean value of the contrast and entropy of each superpixel, and generate a superpixel-based feature image; 采用Otsu阈值检测方法计算所述特征图像的对比度阈值和熵阈值，对海水区域和海冰区域进行分类，并提取所述海冰区域。The Otsu threshold detection method is used to calculate the contrast threshold and the entropy threshold of the feature image, classify the sea water area and the sea ice area, and extract the sea ice area. 2.根据权利要求1所述的方法，其特征在于，所述对所述原始遥感图像的非陆地部分，采用SLIC的超像素分割算法进行超像素分割，包括：2. method according to claim 1, is characterized in that, described to the non-land part of described original remote sensing image, adopts the superpixel segmentation algorithm of SLIC to carry out superpixel segmentation, comprising: 步骤1：设置超像素的边长，生成若干初始超像素种子点；Step 1: Set the side length of the superpixel, and generate several initial superpixel seed points; 步骤2：以一初始超像素种子点为中心，设置对应的临近区域，计算所述邻近区域内每个点的梯度值，将梯度值最小的点作为该临近区域的超像素种子点；Step 2: take an initial superpixel seed point as the center, set the corresponding adjacent area, calculate the gradient value of each point in the adjacent area, and use the point with the smallest gradient value as the superpixel seed point of the adjacent area; 步骤3：设置像素范围，以每个超像素种子点为中心计算所述像素范围内的各个像素点到当前超像素种子点的特征距离；如果当前像素点到当前超像素种子点的特征距离小于其至所属超像素种子点的特征距离，则将当前像素点的所属关系调整至当前超像素种子点；Step 3: Set the pixel range, and calculate the characteristic distance from each pixel point in the pixel range to the current superpixel seed point with each superpixel seed point as the center; if the characteristic distance from the current pixel point to the current superpixel seed point is less than its characteristic distance to the superpixel seed point to which it belongs, then adjust the affiliation of the current pixel point to the current superpixel seed point; 步骤4：对于每个调整后的超像素种子点，计算其超像素中心，作为对应超像素的种子点；Step 4: For each adjusted superpixel seed point, calculate its superpixel center as the seed point of the corresponding superpixel; 步骤5：对步骤3和步骤4进行迭代，直至达到预设迭代次数。Step 5: Iterate steps 3 and 4 until the preset number of iterations is reached. 3.根据权利要求2所述的方法，其特征在于，所述各个像素点到当前超像素种子点的特征距离为：3. method according to claim 2, is characterized in that, the characteristic distance of described each pixel point to current superpixel seed point is:
Figure FDA0002707156240000011
Figure FDA0002707156240000011
 其中，Ni为当前超像素种子点的灰度值；Nj为当前像素点的灰度值；Xi为当前超像素种子点的横坐标；Yi为当前超像素种子点的纵坐标；Xj为当前像素点的横坐标；Yj为当前像素点的纵坐标。Wherein, N i is the grayscale value of the current superpixel seed point; Nj is the grayscale value of the current pixel point; X i is the abscissa of the current superpixel seed point; Y i is the ordinate of the current superpixel seed point; X j is the abscissa of the current pixel; Y j is the ordinate of the current pixel. 4.根据权利要求1所述的方法，其特征在于，所述计算每个超像素的对比度和熵的均值，生成基于超像素的特征图像，包括：4. method according to claim 1, is characterized in that, described calculating the mean value of the contrast of each superpixel and entropy, generates the characteristic image based on superpixel, comprising: 对原始遥感图像的灰度级进行压缩，并设置第一滑动窗口计算灰度共生矩阵；Compress the gray level of the original remote sensing image, and set the first sliding window to calculate the gray level co-occurrence matrix; 根据每个像素的水平灰度共生矩阵，计算像素的对比度和熵；According to the horizontal gray level co-occurrence matrix of each pixel, the contrast and entropy of the pixel are calculated; 计算每个超像素内多个像素点的对比度和熵的均值，作为当前超像素的对比度特征值和熵特征值，生成基于超像素的特征图像。Calculate the average of the contrast and entropy of multiple pixels in each superpixel as the contrast feature value and entropy feature value of the current superpixel, and generate a superpixel-based feature image. 5.根据权利要求4所述的方法，其特征在于，所述像素的对比度和熵为：5. The method according to claim 4, wherein the contrast and entropy of the pixel are:
Figure FDA0002707156240000021
Figure FDA0002707156240000021
Figure FDA0002707156240000022
Figure FDA0002707156240000022
 其中，CON为像素的对比度；ENT为像素的熵；K为第一滑动窗口大小；G(i,j)为灰度共生矩阵第i行j列的值。Among them, CON is the contrast of the pixel; ENT is the entropy of the pixel; K is the size of the first sliding window; G(i, j) is the value of the ith row and j column of the gray level co-occurrence matrix. 6.根据权利要求1所述的方法，其特征在于，所述对海水区域和海冰区域进行分类，包括：6. The method according to claim 1, wherein the classifying the seawater area and the sea ice area comprises: 若超像素的对比度值大于所述对比度阈值和/或超像素的熵值大于所述熵阈值，则所述超像素所在区域为海冰区域；否则为海水区域。If the contrast value of the superpixel is greater than the contrast threshold and/or the entropy value of the superpixel is greater than the entropy threshold, the region where the superpixel is located is a sea ice region; otherwise, it is a seawater region. 7.根据权利要求1所述的方法，其特征在于，还包括：7. The method of claim 1, further comprising: 设置第二滑动窗口，计算所述海水区域和海冰区域在所述第二滑动窗口内的海冰比例，生成海冰密集度特征图像；Setting a second sliding window, calculating the sea ice ratio of the seawater area and the sea ice area in the second sliding window, and generating a sea ice density feature image; 将所述海水区域中的最大连通区域的海冰密集度设置为0；Setting the sea ice concentration of the largest connected area in the seawater area to 0; 将所述海冰密集度特征图像中，海冰密集度大于0的区域识别为海冰，并将海冰识别结果作为海冰提取结果。In the sea ice density feature image, an area with a sea ice density greater than 0 is identified as sea ice, and the sea ice identification result is used as the sea ice extraction result. 8.一种基于超像素分割的海冰提取装置，其特征在于，包括：8. A sea ice extraction device based on superpixel segmentation, characterized in that, comprising: 分离模块，用于对原始遥感图像中的陆地部分和非陆地部分进行分离，得到原始遥感图像的非陆地部分；The separation module is used to separate the land part and the non-land part in the original remote sensing image to obtain the non-land part of the original remote sensing image; 超像素分割模块，用于对所述原始遥感图像的非陆地部分，采用SLIC的超像素分割算法进行超像素分割；A superpixel segmentation module, which is used to perform superpixel segmentation using the superpixel segmentation algorithm of SLIC on the non-land portion of the original remote sensing image; 特征图像生成模块，用于根据超像素分割结果，基于灰度共生矩阵，计算每个超像素的对比度和熵的均值，生成基于超像素的特征图像；The feature image generation module is used to calculate the average value of the contrast and entropy of each superpixel based on the grayscale co-occurrence matrix according to the superpixel segmentation result, and generate a superpixel-based feature image; 分类提取模块，用于采用Otsu阈值检测方法计算所述特征图像的对比度阈值和熵阈值，对海水区域和海冰区域进行分类，并提取所述海冰区域。The classification and extraction module is used for calculating the contrast threshold and the entropy threshold of the feature image by adopting the Otsu threshold detection method, classifying the sea water area and the sea ice area, and extracting the sea ice area. 9.一种电子设备，包括存储器和处理器，所述存储器上存储有计算机程序，其特征在于，所述处理器执行所述程序时实现如权利要求1～7中任一项所述的方法。9 . An electronic device, comprising a memory and a processor, wherein a computer program is stored on the memory, wherein the processor implements the method according to any one of claims 1 to 7 when the processor executes the program . 10.一种计算机可读存储介质，其上存储有计算机程序，其特征在于，所述程序被处理器执行时实现如权利要求1～7中任一项所述的方法。10 . A computer-readable storage medium on which a computer program is stored, characterized in that, when the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.
CN202011042927.3A 2020-09-28 2020-09-28 A method, device and device for sea ice extraction based on superpixel segmentation Pending CN112215104A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011042927.3A CN112215104A (en) 2020-09-28 2020-09-28 A method, device and device for sea ice extraction based on superpixel segmentation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011042927.3A CN112215104A (en) 2020-09-28 2020-09-28 A method, device and device for sea ice extraction based on superpixel segmentation

Publications (1)

Publication Number Publication Date
CN112215104A true CN112215104A (en) 2021-01-12

Family

ID=74051380

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011042927.3A Pending CN112215104A (en) 2020-09-28 2020-09-28 A method, device and device for sea ice extraction based on superpixel segmentation

Country Status (1)

Country Link
CN (1) CN112215104A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114838711A (en) * 2022-03-16 2022-08-02 东北大学秦皇岛分校 Sea ice comprehensive monitoring system and method
CN115147733A (en) * 2022-09-05 2022-10-04 山东东盛澜渔业有限公司 Artificial intelligence-based marine garbage recognition and recovery method
CN115861341A (en) * 2022-11-29 2023-03-28 电子科技大学 Frozen river detection method and device based on mixed second-order gradient matrix eigenvalue
CN117152490A (en) * 2023-08-04 2023-12-01 安徽正弦空间科学技术有限公司 A method, system, storage medium and electronic device for classifying sea ice and sea water
CN120656132A (en) * 2025-08-19 2025-09-16 中国电建集团西北勘测设计研究院有限公司 Marine photovoltaic pile Zhou Fubing early warning method and device, storage medium and electronic equipment

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103839077A (en) * 2014-02-26 2014-06-04 西安电子科技大学 Low-rank-represented polarization SAR image classification method based on superpixel features
CN107170006A (en) * 2017-06-01 2017-09-15 中国科学院遥感与数字地球研究所 The separation method and device of sea ice and Seawater Information in diameter radar image
CN108428220A (en) * 2018-03-05 2018-08-21 武汉大学 Satellite sequence remote sensing image sea island reef region automatic geometric correction method
CN109145718A (en) * 2018-07-04 2019-01-04 国交空间信息技术（北京）有限公司 The road network extracting method and device of remote sensing image based on topology ambiguity
US20200034972A1 (en) * 2018-07-25 2020-01-30 Boe Technology Group Co., Ltd. Image segmentation method and device, computer device and non-volatile storage medium
CN111709483A (en) * 2020-06-18 2020-09-25 山东财经大学 A multi-feature-based superpixel clustering method and device

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103839077A (en) * 2014-02-26 2014-06-04 西安电子科技大学 Low-rank-represented polarization SAR image classification method based on superpixel features
CN107170006A (en) * 2017-06-01 2017-09-15 中国科学院遥感与数字地球研究所 The separation method and device of sea ice and Seawater Information in diameter radar image
CN108428220A (en) * 2018-03-05 2018-08-21 武汉大学 Satellite sequence remote sensing image sea island reef region automatic geometric correction method
CN109145718A (en) * 2018-07-04 2019-01-04 国交空间信息技术（北京）有限公司 The road network extracting method and device of remote sensing image based on topology ambiguity
US20200034972A1 (en) * 2018-07-25 2020-01-30 Boe Technology Group Co., Ltd. Image segmentation method and device, computer device and non-volatile storage medium
CN111709483A (en) * 2020-06-18 2020-09-25 山东财经大学 A multi-feature-based superpixel clustering method and device

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114838711A (en) * 2022-03-16 2022-08-02 东北大学秦皇岛分校 Sea ice comprehensive monitoring system and method
CN115147733A (en) * 2022-09-05 2022-10-04 山东东盛澜渔业有限公司 Artificial intelligence-based marine garbage recognition and recovery method
CN115147733B (en) * 2022-09-05 2022-11-25 山东东盛澜渔业有限公司 Artificial intelligence-based marine garbage recognition and recovery method
CN115861341A (en) * 2022-11-29 2023-03-28 电子科技大学 Frozen river detection method and device based on mixed second-order gradient matrix eigenvalue
CN117152490A (en) * 2023-08-04 2023-12-01 安徽正弦空间科学技术有限公司 A method, system, storage medium and electronic device for classifying sea ice and sea water
CN117152490B (en) * 2023-08-04 2025-08-22 安徽正弦空间科学技术有限公司 Sea ice and seawater classification method, system, storage medium and electronic device
CN120656132A (en) * 2025-08-19 2025-09-16 中国电建集团西北勘测设计研究院有限公司 Marine photovoltaic pile Zhou Fubing early warning method and device, storage medium and electronic equipment

Similar Documents

Publication Publication Date Title
CN112215104A (en) A method, device and device for sea ice extraction based on superpixel segmentation
CN114972329B (en) Image enhancement method and system of surface defect detector based on image processing
CN113781402B (en) Method and device for detecting scratch defects on chip surface and computer equipment
CN107330979B (en) Vector diagram generation method and device for building house type and terminal
CN113781406B (en) Scratch detection method and device for electronic component and computer equipment
CN113034387B (en) Image denoising method, device, equipment and medium
CN107016677A (en) A kind of cloud atlas dividing method based on FCN and CNN
CN115641327B (en) Building engineering quality supervision and early warning system based on big data
CN114332650B (en) Remote sensing image road identification method and system
CN118941582B (en) Geological survey remote sensing image optimization segmentation method
CN112749673A (en) Method and device for intelligently extracting stock of oil storage tank based on remote sensing image
CN106778661A (en) A kind of express lane line detecting method based on morphological transformation and adaptive threshold
CN112329677A (en) Remote sensing image river target detection method and device based on feature fusion
CN117253150A (en) Ship contour extraction method and system based on high-resolution remote sensing image
CN106600613A (en) Embedded GPU-based improved LBP infrared target detection method
CN116071658B (en) SAR image small target detection and recognition method and device based on deep learning
CN108830864A (en) Image partition method
CN114742849A (en) Leveling instrument distance measuring method based on image enhancement
CN119206526B (en) A cloud detection method based on optical remote sensing images combining spectral and texture features
CN117541983B (en) Model data quality analysis method and system based on machine vision
CN108416815B (en) Method and apparatus for measuring atmospheric light value and computer readable storage medium
CN117292133A (en) Super-pixel segmentation method and device for natural image
CN113139975B (en) Road feature-based pavement segmentation method and device
CN119850951B (en) A remote sensing image aircraft segmentation method and system based on aggregation control module
CN119068198B (en) Equipment ice-covered image segmentation and model training method, electronic equipment and storage medium

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
RJ01 Rejection of invention patent application after publication

Application publication date: 20210112

RJ01 Rejection of invention patent application after publication