CN104008366A - 3D intelligent recognition method and system for biology - Google Patents
3D intelligent recognition method and system for biology Download PDFInfo
- Publication number
- CN104008366A CN104008366A CN201410154911.XA CN201410154911A CN104008366A CN 104008366 A CN104008366 A CN 104008366A CN 201410154911 A CN201410154911 A CN 201410154911A CN 104008366 A CN104008366 A CN 104008366A
- Authority
- CN
- China
- Prior art keywords
- dimensional
- unit
- scanning
- identified
- organism
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000001514 detection method Methods 0.000 claims abstract description 23
- 238000012549 training Methods 0.000 claims abstract description 5
- 238000005259 measurement Methods 0.000 claims description 16
- 239000000523 sample Substances 0.000 claims description 14
- 239000012472 biological sample Substances 0.000 claims description 6
- 238000005286 illumination Methods 0.000 claims description 5
- 230000009191 jumping Effects 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 6
- 238000012545 processing Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 4
- 238000003384 imaging method Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 230000004927 fusion Effects 0.000 description 3
- 238000000513 principal component analysis Methods 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000003321 amplification Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012850 discrimination method Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000002310 reflectometry Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004611 spectroscopical analysis Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 239000013598 vector Substances 0.000 description 1
Landscapes
- Image Processing (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a 3D intelligent recognition method and system for biology. The method includes the steps that S1, a biology standard sample 3D image database is obtained through training; S2, 3D images of biology to be recognized are obtained; S3, the 3D images of the biology to be recognized are compared with the biology standard sample 3D image database for recognition, whether the similarity reaches a preset threshold value or not is judged, if yes, S4 is executed, and if not, the method is skipped to S2; S4, the 3D images are marked and output. The system comprises a 3D scanning device, a control device connected with the 3D scanning device, an irradiation device, a camera shooting device, a position detection unit and a scanning area determining unit, wherein the irradiation device, the camera shooting device, the position detection unit and the scanning area determining unit are respectively connected with the control device. The 3D images of the biology to be recognized are obtained and compared with the biology standard sample 3D image database for recognition, whether the similarity reaches the preset threshold value or not is judged, the 3D images are marked and output, the whole process is operated completely in a full-automatic mode, the biology is recognized intelligently in a full-automatic mode, and recognition efficiency and registration accuracy are remarkably improved.
Description
Technical Field
The invention relates to the technical field of three-dimensional scanning, in particular to a 3D biological intelligent identification method and a system.
Background
Three-dimensional scanning is a high and new technology integrating light, mechanical, electrical and computer technologies, and is mainly used for scanning the spatial appearance, structure and color of an object to obtain the spatial coordinates of the surface of the object. The three-dimensional space information of the object is converted into a digital signal which can be directly processed by a computer, and a very convenient and fast means is provided for object digitization.
Common three-dimensional scanning equipment is divided into a contact type and a non-contact type according to different sensing modes. In the contact type, a detection head is directly contacted with the surface of an object, and a photoelectric signal fed back by the detection head is converted into digital surface shape information, so that the scanning and the measurement of the surface shape of the object are realized, and a three-coordinate measuring machine is mainly taken as a representative. The contact type measurement has higher accuracy and reliability, and can quickly and accurately measure the basic geometric shapes of objects such as planes, circles, cylinders, cones, spheres and the like by matching with measurement software. Therefore, the radius of the probe needs to be compensated to obtain the real shape of the object, so that the problem of correction error can be caused; when the contact probe is used for measurement, the force of the contact probe can cause the local deformation between the probe tip part and a measured object to influence the actual reading of the measured value; the probe is overrun due to inertia and time delay of the probe trigger mechanism, and dynamic error is generated due to approaching speed.
The non-contact photoelectric method has become a great trend for rapidly measuring the three-dimensional shape of a curved surface, the non-contact measurement not only avoids the trouble caused by compensating the radius of a measuring head in the contact measurement, but also can realize high-speed three-dimensional scanning on various surfaces. At present, there are many non-contact three-dimensional scanners, and according to different sensing methods, there are commonly used three-dimensional laser scanners, photographic three-dimensional scanners, CT scanners and the like which respectively represent the mainstream in the market based on laser scanning measurement, structured light scanning measurement, industrial CT and the like. The non-contact three-dimensional scanner does not damage the surface of an object due to non-direct contact, and has the characteristics of high speed, easiness in operation and the like compared with a contact type scanner, the three-dimensional laser scanner can reach the speed of 5000 plus 10000 points/second, and the photographic three-dimensional scanner adopts surface light, so that the speed reaches several seconds and millions of measuring points, and the photographic three-dimensional scanner is applied to real-time scanning and has good advantages in industrial detection.
The measurement research of three-dimensional images by a photoelectric method is increasingly paid attention by people, but the defects that the mechanical structure is complex, the measurement range is limited by the size of a mechanical device, the speed is low due to point-to-point measurement and the like still exist. Meanwhile, the photoelectric method for measuring the three-dimensional surface shape is influenced by the unevenness of the surface characteristics (such as height mutation, shadow and insufficient reflectivity) of the object, and the like, so that the optical signal processing is difficult.
Disclosure of Invention
The invention aims to solve the technical problems of complex three-dimensional scanning structure, limited measuring range, low speed and the like in the prior art, and provides a 3D biological intelligent identification method and a system.
The technical scheme adopted by the invention for solving the technical problems is as follows:
according to an aspect of the present invention, a 3D biological intelligent recognition method is provided, which specifically includes the following steps:
s1: training to obtain a biological sample three-dimensional image database, wherein the biological sample three-dimensional image is a biological three-dimensional image containing three-dimensional data, and the three-dimensional data is n preset number of characteristic values (x, y, z);
s2: acquiring a biological three-dimensional image containing three-dimensional data of a biological to be identified;
s3: comparing and recognizing the three-dimensional image of the organism to be recognized with the three-dimensional image database of the biological standard sample, judging whether the similarity reaches a preset threshold value, and if so, executing the step S4; otherwise, jumping to step S2;
s4: and identifying and outputting the three-dimensional image of the living being to be identified, which reaches a preset threshold value.
Preferably, the step of acquiring a three-dimensional image containing three-dimensional data comprises:
s11: shooting at least two groups of n pictures with preset number by using at least two cameras or 3D cameras; the creature comprises n feature points with preset quantity, the feature points of each photo are different, and each feature point corresponds to a different feature value (x, y, z);
s12: acquiring three-dimensional data of at least two groups of pictures;
s13: the three-dimensional data of at least two groups of each picture are registered and fused into a three-dimensional image of the organism.
Preferably, the z value among the characteristic values (x, y, z) is calculated by a small angle difference frequency method.
Preferably, the acquiring three-dimensional data of at least two groups of each photo can also be:
s121: directly acquiring characteristic values (x, y) of characteristic points of at least two groups of pictures;
s122: finding the closest matching z0 value in the standard library according to the characteristic values (x, y) of the characteristic points of at least two groups of photos;
s123: and combining the acquired characteristic values (x, y) with the z0 value in the standard library to form three-dimensional data.
According to another aspect of the present invention, a 3D biological intelligent recognition system is provided, which includes a three-dimensional scanning device for acquiring three-dimensional data of a living being to be recognized, a control device connected to the three-dimensional scanning device for registering and fusing the three-dimensional data to obtain a three-dimensional image of the living being to be recognized, and an irradiation device, an image pickup device, a position detection unit, and a scanning area determination unit respectively connected to the control device; wherein,
the irradiation device is used for emitting a gap-shaped light beam to the organism to be identified based on the control signal sent by the control device;
the camera device is used for sequentially shooting the photos of the organisms to be identified irradiated by the slit-shaped light beams;
a position detection unit for detecting a position of the slit-shaped light beam in the picture by scanning the picture taken by the image pickup device;
a scanning area determination unit for determining a scanning area of the position detection unit in the photograph as the living organism to be recognized based on a position of the slit-like light beam in the photograph taken by the image pickup device before the photograph as the living organism to be recognized.
Preferably, the control device includes: a complete scanning data acquisition unit and a three-dimensional model acquisition unit; wherein,
the complete scanning data acquisition unit is used for registering and fusing at least two groups of three-dimensional data to obtain complete scanning data of the side, the top and the bottom of the organism to be identified;
and the three-dimensional model acquisition unit is connected with the complete scanning data acquisition unit and is used for registering and fusing the complete scanning data of the side, the top and the bottom of the organism to be identified to obtain a three-dimensional image of the organism to be identified.
Preferably, the irradiation device includes: a light emitting unit and a reflecting mirror; wherein,
the light-emitting unit is used for emitting a gap-shaped light beam to the reflecting surface of the reflector based on a first control signal emitted by the control device; the first control signal is a signal for controlling whether the light-emitting unit emits the gap-shaped light beam or not;
the reflecting mirror is used for controlling the rotation of the reflecting surface of the reflecting mirror based on a second control signal sent by the control device and reflecting the gap-shaped light beam to the creature to be identified; the second control signal is a signal for controlling the rotation angle of the reflecting surface of the mirror.
Preferably, the control device further includes: an irradiation control unit, an image information acquisition unit, and a shape measurement unit; wherein,
the irradiation control unit is used for sending the first control signal to the light-emitting unit and sending a second control signal to the reflector;
an image information acquisition unit for selectively reading out and obtaining information on pixel luminance as image information from photodetection corresponding to the scanning area input by the scanning area determination unit captured by the image pickup device;
a shape measuring unit for measuring information of a three-dimensional shape of a living being to be recognized as shape information.
Preferably, the position detection unit is further configured to scan the pixels in the scanning area determined by the scanning area determination unit and detect a position of a brightest pixel having a pixel luminance exceeding a preset threshold in the photograph, and take the position as the light beam position information in the photograph.
One of the above technical solutions of the present invention has the following advantages or beneficial effects: the method comprises the steps of obtaining a three-dimensional image of a living being to be identified, comparing and identifying the three-dimensional image with a living being sample three-dimensional image database, judging whether the three-dimensional image reaches a preset threshold value, and identifying and outputting the three-dimensional image reaching the preset threshold value. The whole process is fully-automatic, human participation is not needed, full-automatic intelligent identification of the organisms to be identified is realized, and the identification efficiency and the registration precision are both obviously improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without inventive efforts, wherein:
FIG. 1 is a schematic flow chart of an embodiment of a 3D biological intelligent recognition method of the invention;
FIG. 2 is a schematic structural diagram of a first embodiment of a 3D biological intelligent recognition system according to the present invention;
FIG. 3 is a schematic structural diagram of a second embodiment of the 3D biological intelligent recognition system of the present invention;
fig. 4 is a schematic diagram of a camera system of an embodiment of the 3D biometric identification system of the present invention.
Fig. 5 is a schematic diagram of the application of the 3D biological intelligent recognition system using the CCD camera according to the present invention.
Detailed Description
In order that the objects, aspects and advantages of the invention will become more apparent, the various embodiments described hereinafter refer to the accompanying drawings which form a part hereof, and in which are shown by way of illustration various embodiments in which the invention may be practiced. It is to be understood that other embodiments may be utilized and structural and functional modifications may be made to the embodiments set forth herein without departing from the scope and spirit of the present invention.
The embodiment of the 3D biological intelligent identification method disclosed by the invention comprises the following steps as shown in figure 1:
s1: training to obtain a biological sample three-dimensional image database, wherein the biological sample three-dimensional image is a biological three-dimensional image containing three-dimensional data; wherein the three-dimensional data is a predetermined number n of eigenvalues (x, y, z);
the biological three-dimensional image containing three-dimensional data is acquired in the following two ways:
the first mode is to adopt at least two cameras or 3D cameras, and the specific steps are as follows:
s11: shooting at least two groups of n pictures with preset number by using at least two cameras or 3D cameras;
s12: acquiring three-dimensional data of at least two groups of pictures;
s13: the three-dimensional data of at least two groups of each picture are registered and fused into a three-dimensional image of the organism.
In this embodiment, the z value in step S12 can be directly calculated by a small-angle difference frequency method, and can also be obtained by the following method:
s121: directly acquiring characteristic values (x, y) of characteristic points of at least two groups of pictures;
s122: finding the closest matching z0 value in the standard library according to the characteristic values (x, y) of the characteristic points of at least two groups of photos;
s123: and combining the acquired characteristic values (x, y) with the z0 value in the standard library to form three-dimensional data.
Specifically, the obtained characteristic value (x, y) and the z0 value in a standard library are combined into three-dimensional data by directly obtaining the characteristic value (x, y) from the photos as the coordinate value of the characteristic point, then calculating the characteristic value z by using a small angle difference frequency method or other methods, and further finding the most matched and closest z0 value in the standard library according to the characteristic value (x, y) of the characteristic point of each photo; and compared to a predetermined threshold (e.g., 95% or 98%), and if above the predetermined threshold, the actual z value is used to form the three-dimensional data using z0 for the standard library, and if below the predetermined threshold.
Presetting n feature points with preset quantity for each living being, wherein the feature points of each photo are different, and each feature point corresponds to different feature values (x, y, z); in order to conveniently establish a three-dimensional image, a preset number n of feature points are distributed in a plurality of areas of the photo, and the areas are used for dividing the photo into an upper area, a middle area and a lower area. The n feature points of the predetermined number may be in the three regions, and may be distributed more evenly or may be distributed more in a certain region.
The second mode is to adopt a single camera, and the specific steps are as follows:
s11': the method comprises the steps of shooting n biological photos of a predetermined number of organisms, and acquiring feature data of feature points of each photo as feature values (x, y) of the photo; the characteristic value (x, y) is a coordinate value of a characteristic point, and the coordinate value of the specific point can be directly obtained from the photo;
s12': according to the shot picture, finding the z value which is most closely matched in the standard library as the characteristic value z of the characteristic point; the standard library comprises various typical biological models, mainly comprises z values of different styles or modes, finds the closest matched z value as the characteristic z value of the characteristic point, and obtains the complete characteristic values (x, y, z) of n characteristic points; the z-values in the standard library were determined as: similar organisms are obtained by sampling (e.g., 100 or more), by probabilistic methods, their z-values are measured, and averaged to obtain the average z-value in such organisms as the z-value of the standard library.
S13': the three-dimensional data of each photograph is registered and fused into a three-dimensional image of the organism.
S2: acquiring a biological three-dimensional image containing three-dimensional data of a biological to be identified; the specific acquisition manner is the same as step S1.
S3: comparing and identifying the three-dimensional image of the living being in the S2 with a living being sample three-dimensional image database, judging whether the similarity reaches a preset threshold value, and if so, performing the step S4; otherwise, jumping to step S2; wherein the predetermined threshold is 95%, and even 98% can be set.
S4: a three-dimensional image of the living being reaching a predetermined threshold is identified and output.
In order to more clearly describe the 3D bio-intelligent recognition method of the present invention, the following description will be made in detail by taking specific face recognition as a case. The three regions above the face are scanned, and 64 feature points are found in the three regions, wherein the feature points can be distributed more uniformly or a certain region is distributed more. Taking 64 points as an example, or more or less points, and adopting a single camera, the specific steps are as follows:
(1) training to obtain biological standard sample three-dimensional image database
A. Firstly, shooting 64 photos of a human face by using a camera, wherein feature data of one feature point is recorded or obtained in each photo, the feature data comprises 64 feature points, 64 photos and feature values (x, y and z) of the 64 feature points, and the coordinate values of x and y can be directly obtained from the photos;
B. the z-coordinate values are obtained by analog approximation: according to the shot pictures, finding the closest matched z value in a standard library (containing various typical face models and mainly having different styles or patterned z values) to be used as the characteristic value z of the characteristic point, thereby obtaining the complete characteristic values (x, y, z) of 64 points;
C. and combining the feature values (x, y, z) of the 64 feature points into a biological standard sample three-dimensional image of the human face, and storing the biological standard sample three-dimensional image database.
(2) And shooting a human face three-dimensional image containing three-dimensional data of the human face by using a camera. That is, when the recognition is applied, the feature values (x, y, z) of 64 feature points of the actual face are also obtained according to the above steps.
A. Firstly, shooting 64 photos of a human face by using a camera, wherein feature data of one feature point is recorded or obtained in each photo, the feature data comprises 64 feature points, 64 photos and feature values (x, y and z) of the 64 feature points, and the coordinate values of x and y can be directly obtained from the photos;
B. the z-coordinate values are obtained by analog approximation: according to the shot pictures, finding the closest matched z value in a standard library (containing various typical face models and mainly having different styles or patterned z values) to be used as the characteristic value z of the characteristic point, thereby obtaining the complete characteristic values (x, y, z) of 64 points;
C. and combining the feature values (x, y, z) of the 64 feature points into a three-dimensional image of the human face.
(3) Comparing and recognizing the acquired three-dimensional image with a biological standard sample three-dimensional image database, judging whether the similarity reaches a preset threshold value, and if not, returning to the step (2) to shoot the face image again; and if so, identifying and outputting the three-dimensional image reaching the preset threshold value.
Based on the analysis of the model for imaging the surface of a living being by means of spectroscopy, the types of biological local differences between near infrared and visible light are summarized. Based on the characteristics, a process of filtering, local mode coding, feature extraction and feature selection is provided to obtain the features with the biological invariance and the discrimination capability of near infrared and visible light. Two image local modes are proposed, namely an edge mode and a local hidden edge mode. Finally, the effectiveness of the algorithm is shown in the experiment of near infrared and visible light biological cross recognition.
(1) The target identification algorithm based on visible light image template matching has small calculation amount and easy parallel implementation, and only takes 10ms to process a frame of real-time image on a VTS-642 DSP system.
As shown in fig. 2, the correlation matching algorithm based on visible light image template matching is divided into two steps, first, a reference template of a target is prepared by using a visible light image, and the prepared reference template is subjected to quantization to generate a quantized reference image; then preprocessing the real-time image such as resolution reduction, homogeneous transformation, OSTI clustering and the like; and traversing and searching in the processed real-time image by using the structural template, searching for the correlation between the reference image and the real-time image, and positioning by taking the place with the strongest correlation as a daily mark area.
(2) The elastic image matching biological recognition algorithm based on the characteristics is improved, and the improved algorithm has high recognition rate and good real-time performance
Aiming at the problems of high space complexity and poor real-time performance of the traditional biometric elastic image matching algorithm, the elastic image matching improvement algorithm is provided, after the characteristic points of a biometric image are preprocessed by Gabor wavelet, the generated characteristic vectors are processed by combining Principal Component Analysis (PCA) and a Fisher linear discrimination method (PLD), the dimension is reduced, the calculated amount is reduced, and meanwhile, the recognition speed is improved on the premise of not reducing the recognition rate. Compared with the traditional PCA algorithm, FLD algorithm and EGM algorithm, the improved algorithm is proved to have high recognition rate and good real-time property.
(3) Improve the modeling equation of the 3D biological deformation model and improve the identification precision
In the biological recognition technology, because a single 2D image can not provide complete information required by recognition, the recognition precision is difficult to improve, and in the biological recognition process, feature extraction is an important link influencing the recognition effect.
According to the 3D biological intelligent identification method, through actual test, the actual biological can be accurately identified within 15 degrees of left-right skew, when the method is used, relevant data obtained after the actual biological is shot by a camera is compared with a database obtained by sampling, and the overlapping degree is observed, so that the coincidence is 95 percent or even 98 percent.
As shown in fig. 3, the embodiment of the 3D biological intelligent recognition system of the present invention includes a three-dimensional scanning device 6 for acquiring at least two sets of three-dimensional data of a living being to be recognized, a control device 5 connected to the three-dimensional scanning device 6 for registering and fusing the at least two sets of three-dimensional data to obtain a three-dimensional image of the living being to be recognized, and an irradiation device 1, an image pickup device 2, a position detection unit 3, a scanning area determination unit 4, and a storage unit 7 respectively connected to the control device 5.
The illuminating device 1 is used for emitting a gap-shaped light beam to the organism to be identified based on a control signal sent by the control device 5, and when the organism to be identified changes the image pick-up position, the illuminating device 1 can emit the gap-shaped light beam while changing the image pick-up position relative to the organism to be identified; the camera device 2 is used for sequentially shooting the photos of the organisms to be identified irradiated by the slit-shaped light beams; a position detection unit 3 for detecting the position of the slit-like light beam in the photograph by scanning the photograph taken by the image pickup device 2; a scanning-area determining unit 4 for determining a scanning area of the position detecting unit 3 in the photograph as the living organism to be recognized based on the position of the slit-like light beam in the photograph taken by the image pickup device 2 before the photograph as the living organism to be recognized. The storage unit 7 is used to store image information, beam position information, shape information, and the like. The image information is information representing an image captured by the imaging device 2, the light beam position information is information representing the position of the light beam in each image captured by the imaging device 2, and the shape information is information representing the three-dimensional shape of the living organism to be recognized measured by the 3D biometric identification system.
Among them, the position detection unit 3 is also configured to detect the light beam position in each photograph sequentially taken by the image pickup device 2 based on the image information read out from the storage unit 7, specifically, the position detection unit 3 scans pixels in the scanning area determined by the scanning area determination unit 4, and then, the position detection unit 3 detects the position in the image of the brightest pixel among the pixels whose luminance exceeds a predetermined threshold among the scanned pixels, as the light beam position of the photograph, outputs the detection result to the shape measurement unit 55 and stores in the storage unit 7 as the light beam position information.
In the embodiment of the 3D biological intelligent recognition system of the present invention, as shown in fig. 4, the illumination device 1 includes a light emitting unit 11 and a reflector 12; the light emitting unit 11 is used for emitting a gap-shaped light beam to the reflecting surface of the reflector 12 based on a first control signal emitted by the control device 5; the first control signal is a signal for controlling whether the light emitting unit 11 emits the gap-shaped light beam; a mirror 12 for controlling the rotation of the reflecting surface of the mirror 12 and reflecting the slit-shaped light beam to the living being to be identified based on a second control signal issued by the control device 5; the second control signal is a signal for controlling the rotation angle of the reflecting surface of the mirror 12. Among them, the reflecting mirror 12 is a reflecting plate: by controlling the rotation of the reflection surface based on a control signal issued from the control unit 5, the light beam emitted from the light emitting unit 11 is emitted so that the irradiation position on the platform and the living organism to be identified is moved from the negative direction of the X axis to the positive direction.
Specifically, the 3D biometric intelligent recognition system controls the light emitting unit 11 to emit a light beam to the living being to be recognized while changing the irradiation position with respect to the living being to be recognized by rotating the reflecting mirror 12. The 3D biological intelligent recognition system emits a light beam from an obliquely upward direction with respect to the living being to be recognized while moving an irradiation position of the light beam on the living being to be recognized from a negative direction of the X-axis toward a positive direction. Next, the 3D biometric identification system sequentially captures images of the living being to be identified irradiated with the light beam by the image pickup device 2. That is, the image pickup device 2 picks up an image of the light beam moving in the moving direction for each image pickup frame on the platform and the living organism to be recognized by receiving the light beams reflected by the platform and the living organism to be recognized. Next, the position detection unit 3 detects the position of the light beam in each image captured by the imaging device 2, and measures the three-dimensional shape of the living organism to be recognized according to the principle of triangulation using the detected light beam position, more specifically, information of the three-dimensional shape of the living organism to be recognized is measured as shape information by the shape measurement unit 55.
The 3D biological intelligent recognition system comprises one or three same three-dimensional scanning devices 6, wherein the three same three-dimensional scanning devices are respectively a horizontal scanning device, a top scanning device and a bottom scanning device; the horizontal scanning device is used for acquiring three-dimensional information of the side face of a living being to be identified to obtain a group of horizontally scanned three-dimensional data; the bottom scanning device is used for acquiring the three-dimensional information of the bottom of the organism to be identified to obtain a group of bottom-scanned three-dimensional data; the top scanning device is used for acquiring the three-dimensional information of the top of the creature to be identified and obtaining a group of top-scanned three-dimensional data. Meanwhile, the three-dimensional scanning device 6 is also configured to acquire three-dimensional data of the coordinate conversion reference object under the control of the control device 5.
Wherein when the position detection unit 3 scans the image while sequentially moving the scanning start position in each scanning in the scanning direction in the direction orthogonal to the scanning direction, which is the direction in which the slit-shaped light beam is moved on the image by changing the irradiation position, the scanning region determination unit 4 determines to scan each scanning region in the scanning direction.
The control device 5 of the 3D biological intelligent recognition system of the invention, the control device 5 is a processing unit which generally controls the operation of the whole 3D biological intelligent recognition system, and specifically comprises a complete scanning data acquisition unit 51 and a three-dimensional model acquisition unit 52; the complete scanning data acquisition unit 51 is configured to sequentially register and fuse at least two sets of three-dimensional data to obtain complete scanning data of the side, top, and bottom of the living being to be identified; and the three-dimensional model acquisition unit 52 is connected with the complete scanning data acquisition unit 51 and is used for registering and fusing the complete scanning data of the side, the top and the bottom of the organism to be identified to obtain a three-dimensional image of the organism to be identified.
The control device 5 further includes an irradiation control unit 53, an image information acquisition unit 54, and a shape measurement unit 55; wherein the illumination control unit 53 is configured to send a first control signal to the light emitting unit 11 and a second control signal to the reflector 12, and the illumination control unit 53 outputs information indicating the rotation angle of the reflector 12 to the shape measuring unit 55. The image information obtaining unit 54 is a processing unit that obtains images sequentially captured by the image capturing apparatus 2 from the image capturing apparatus 2 and stores them as image information in the storage unit 7, specifically, the image information obtaining unit 54 that selectively reads out and obtains information on pixel luminance as image information from photodetection corresponding to the scanning area input by the scanning area determining unit 4 captured by the image capturing apparatus 2; a shape measuring unit 55 for measuring information of a three-dimensional shape of the living being to be recognized as the shape information.
As shown in fig. 5, the 3D bio-intelligent recognition system of the present invention employs a CCD camera mainly for measuring energy distribution. The energy distribution of the measured light field is imaged on a CCD area array detector through an optical system; the signal generated by the photoelectric effect of the detector is processed by a camera processing circuit to complete two-dimensional scanning reading, amplification, system conversion and the like, so as to form a standard full television signal; the signal is digitized by the camera processing control board and communicated to the computer for various processing.
The 3D biological intelligent identification system respectively carries out three-dimensional scanning operation from different angles, respectively carries out automatic registration and fusion on the side scanning three-dimensional data set, the bottom scanning data set and the top scanning data set, and then carries out automatic registration and fusion on at least two groups of three-dimensional data with different angles after fusion, thereby obtaining a complete three-dimensional model of the organism to be identified. The whole process is fully-automatic, manual participation is not needed, manual labeling auxiliary registration processing is not needed, full-automatic three-dimensional information acquisition of the organisms to be identified is achieved, and the scanning efficiency and the registration accuracy are remarkably improved.
While the invention has been described with reference to a preferred embodiment, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the spirit and scope of the invention. In addition, many modifications may be made to adapt a particular situation or material to the teachings of the invention without departing from the essential scope thereof. Therefore, it is intended that the invention not be limited to the particular embodiment disclosed, but that the invention will include all embodiments falling within the scope of the appended claims.
Claims (10)
1. A3D biological intelligent identification method is characterized by comprising the following steps:
s1: training to obtain a biological sample three-dimensional image database, wherein the biological sample three-dimensional image is a biological three-dimensional image containing three-dimensional data, and the three-dimensional data is n preset number of characteristic values (x, y, z);
s2: acquiring a biological three-dimensional image containing three-dimensional data of a biological to be identified;
s3: comparing and recognizing the three-dimensional image of the organism to be recognized with the organism sample three-dimensional image database, judging whether the similarity reaches a preset threshold value, and if so, executing a step S4; otherwise, jumping to step S2;
s4: and identifying and outputting the three-dimensional image of the living beings to be recognized reaching the preset threshold value.
2. The 3D biometric intelligent recognition method according to claim 1, wherein the step of obtaining the three-dimensional image containing the three-dimensional data is:
s11: shooting at least two groups of n pictures with preset number by using at least two cameras or 3D cameras;
s12: acquiring three-dimensional data of at least two groups of pictures;
s13: registering and fusing the three-dimensional data of at least two sets of each of the photographs into a three-dimensional image of the living being.
3. The 3D biometric intelligent recognition method according to claim 2, wherein z-value among the feature values (x, y, z) is calculated by a small angle difference frequency method.
4. The 3D biometric intelligent recognition method according to claim 2, wherein the obtaining three-dimensional data of at least two groups of each photo is further:
s121: directly acquiring characteristic values (x, y) of characteristic points of at least two groups of pictures;
s122: finding the closest matching z0 value in a standard library according to the characteristic values (x, y) of the characteristic points of at least two groups of the photos;
s123: and combining the acquired characteristic values (x, y) with the z0 value in the standard library to form three-dimensional data.
5. The 3D biometric method according to any one of claims 2 or 4, wherein the biometric device comprises a predetermined number n of feature points, the feature points of each of the pictures are different, and each of the feature points corresponds to a different feature value (x, y, z).
6. A3D biological intelligent identification system comprises a three-dimensional scanning device (6) used for obtaining three-dimensional data of a to-be-identified organism and a control device (5) connected with the three-dimensional scanning device (6) and used for registering and fusing the three-dimensional data to obtain a three-dimensional image of the to-be-identified organism; the 3D biological intelligent recognition system is characterized by further comprising an irradiation device (1), a camera unit (2), a position detection unit (3) and a scanning area determination unit (4), wherein the irradiation device, the camera unit (2), the position detection unit and the scanning area determination unit are respectively connected with the control device (5);
the irradiation device (1) is used for emitting a gap-shaped light beam to the organism to be identified based on a control signal sent by the control device (5);
the camera shooting unit (2) is used for sequentially shooting the pictures of the organisms to be identified irradiated by the gap-shaped light beams;
the position detection unit (3) is used for detecting the position of the gap-shaped light beam in the picture by scanning the picture shot by the camera shooting unit (2);
the scanning area determination unit (4) is used for determining the scanning area of the position detection unit (3) in the picture as the living being to be identified based on the position of the gap-shaped light beam in the picture taken by the image pickup unit (2) before the picture as the living being to be identified.
7. The 3D biometric intelligent recognition system according to claim 6, wherein the control means (5) comprises: a complete scan data acquisition unit (51) and a three-dimensional model acquisition unit (52);
the complete scanning data acquisition unit (51) is used for registering and fusing the at least two groups of three-dimensional data to obtain complete scanning data of the side, the top and the bottom of the organism to be identified;
the three-dimensional model acquisition unit (52) is connected with the complete scanning data acquisition unit (51) and is used for registering and fusing complete scanning data of the side, the top and the bottom of the organism to be identified to obtain a three-dimensional image of the organism to be identified.
8. The 3D biometric intelligent recognition system according to claim 6, wherein the illumination device (1) comprises: a light emitting unit (11) and a reflector (12);
the light-emitting unit (11) is used for emitting the gap-shaped light beam to the reflecting surface of the reflector (12) based on a first control signal emitted by the control device (5); the first control signal is a signal for controlling whether the light-emitting unit (11) emits the gap-shaped light beam or not;
the reflector (12) is used for controlling the rotation of the reflecting surface of the reflector (12) and reflecting the gap-shaped light beam to the organism to be identified based on a second control signal emitted by the control device (5); the second control signal is a signal for controlling a rotation angle of a reflecting surface of the reflecting mirror (12).
9. The 3D biometric intelligent recognition system according to claim 8, wherein the control device (5) further comprises: an irradiation control unit (53), an image information acquisition unit (54), and a shape measurement unit (55);
-said illumination control unit (53) for issuing said first control signal to said light emitting unit (11) and said second control signal to said mirror (12);
the image information acquisition unit (54) is used for selectively reading out and obtaining information about pixel brightness as image information from photoelectric detection which is shot by the image pickup device (2) and corresponds to the scanning area input by the scanning area determination unit (4);
the shape measurement unit (55) is configured to measure information of a three-dimensional shape of the living organism to be identified as shape information.
10. The 3D bio-intelligent recognition system according to claim 9, wherein the position detection unit (3) is further configured to scan pixels in the scanning area determined by the scanning area determination unit (4) and detect a position of a brightest pixel in the photo where a pixel brightness exceeds a preset threshold, and to take the position as the light beam position information in the photo.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410154911.XA CN104008366A (en) | 2014-04-17 | 2014-04-17 | 3D intelligent recognition method and system for biology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410154911.XA CN104008366A (en) | 2014-04-17 | 2014-04-17 | 3D intelligent recognition method and system for biology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104008366A true CN104008366A (en) | 2014-08-27 |
Family
ID=51369015
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410154911.XA Pending CN104008366A (en) | 2014-04-17 | 2014-04-17 | 3D intelligent recognition method and system for biology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104008366A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104504682A (en) * | 2014-11-25 | 2015-04-08 | 中国农业大学 | Animal shape phenotype automatic measuring method |
| WO2018040099A1 (en) * | 2016-08-31 | 2018-03-08 | 深圳市唯特视科技有限公司 | Three-dimensional face reconstruction method based on grayscale and depth information |
| CN109726611A (en) * | 2017-10-27 | 2019-05-07 | 北京小米移动软件有限公司 | Biological feather recognition method and device, readable storage medium storing program for executing and electronic equipment |
| CN110334691A (en) * | 2019-07-16 | 2019-10-15 | 武汉成名仪器有限公司 | Microorganism water quality monitoring method, system and storage medium |
| CN111212598A (en) * | 2018-07-27 | 2020-05-29 | 合刃科技（深圳）有限公司 | Biological feature recognition method, device and system and terminal equipment |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1404016A (en) * | 2002-10-18 | 2003-03-19 | 清华大学 | Establishing method of human face 3D model by fusing multiple-visual angle and multiple-thread 2D information |
| CN1595280A (en) * | 2004-07-01 | 2005-03-16 | 上海交通大学 | Three-dimensional human face rebuilding system on rotating platform |
| CN101303772A (en) * | 2008-06-20 | 2008-11-12 | 浙江大学 | A Nonlinear 3D Face Modeling Method Based on Single Image |
| CN101980243A (en) * | 2010-10-15 | 2011-02-23 | 中国人民解放军国防科学技术大学 | Method and device for three-dimensional recognition of finger veins based on binocular vision |
| CN102735166A (en) * | 2011-04-14 | 2012-10-17 | 株式会社安川电机 | Three-dimensional scanner and robot system |
| CN103279982A (en) * | 2013-05-24 | 2013-09-04 | 中国科学院自动化研究所 | Robust rapid high-depth-resolution speckle three-dimensional rebuilding method |
| CN103292740A (en) * | 2013-05-24 | 2013-09-11 | 贵阳嘉瑜光电科技咨询中心 | Three-dimensional scanner measuring method and device |
| CN103292697A (en) * | 2013-05-27 | 2013-09-11 | 深圳先进技术研究院 | Three-dimensional scanning system |
| CN103716615A (en) * | 2014-01-09 | 2014-04-09 | 西安电子科技大学 | 2D video three-dimensional method based on sample learning and depth image transmission |
-
2014
- 2014-04-17 CN CN201410154911.XA patent/CN104008366A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1404016A (en) * | 2002-10-18 | 2003-03-19 | 清华大学 | Establishing method of human face 3D model by fusing multiple-visual angle and multiple-thread 2D information |
| CN1595280A (en) * | 2004-07-01 | 2005-03-16 | 上海交通大学 | Three-dimensional human face rebuilding system on rotating platform |
| CN101303772A (en) * | 2008-06-20 | 2008-11-12 | 浙江大学 | A Nonlinear 3D Face Modeling Method Based on Single Image |
| CN101980243A (en) * | 2010-10-15 | 2011-02-23 | 中国人民解放军国防科学技术大学 | Method and device for three-dimensional recognition of finger veins based on binocular vision |
| CN102735166A (en) * | 2011-04-14 | 2012-10-17 | 株式会社安川电机 | Three-dimensional scanner and robot system |
| CN103279982A (en) * | 2013-05-24 | 2013-09-04 | 中国科学院自动化研究所 | Robust rapid high-depth-resolution speckle three-dimensional rebuilding method |
| CN103292740A (en) * | 2013-05-24 | 2013-09-11 | 贵阳嘉瑜光电科技咨询中心 | Three-dimensional scanner measuring method and device |
| CN103292697A (en) * | 2013-05-27 | 2013-09-11 | 深圳先进技术研究院 | Three-dimensional scanning system |
| CN103716615A (en) * | 2014-01-09 | 2014-04-09 | 西安电子科技大学 | 2D video three-dimensional method based on sample learning and depth image transmission |
Non-Patent Citations (3)
| Title |
|---|
| 宋顶利 等: ""关键点匹配三维人脸识别方法"", 《计算机应用研究》 * |
| 程潇: ""关于三维人脸图像数据融合的研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 陈立生 等: ""基于几何特征与深度数据的三维人脸识别"", 《电脑知识与技术》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104504682A (en) * | 2014-11-25 | 2015-04-08 | 中国农业大学 | Animal shape phenotype automatic measuring method |
| WO2018040099A1 (en) * | 2016-08-31 | 2018-03-08 | 深圳市唯特视科技有限公司 | Three-dimensional face reconstruction method based on grayscale and depth information |
| CN109726611A (en) * | 2017-10-27 | 2019-05-07 | 北京小米移动软件有限公司 | Biological feather recognition method and device, readable storage medium storing program for executing and electronic equipment |
| CN109726611B (en) * | 2017-10-27 | 2021-07-23 | 北京小米移动软件有限公司 | Biological feature recognition method and device, readable storage medium and electronic equipment |
| CN111212598A (en) * | 2018-07-27 | 2020-05-29 | 合刃科技（深圳）有限公司 | Biological feature recognition method, device and system and terminal equipment |
| CN111212598B (en) * | 2018-07-27 | 2023-09-12 | 合刃科技（深圳）有限公司 | Biological feature recognition method, device, system and terminal equipment |
| CN110334691A (en) * | 2019-07-16 | 2019-10-15 | 武汉成名仪器有限公司 | Microorganism water quality monitoring method, system and storage medium |
| CN110334691B (en) * | 2019-07-16 | 2021-10-22 | 武汉成名仪器有限公司 | Microorganism water quality monitoring method, system and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US6549288B1 (en) | Structured-light, triangulation-based three-dimensional digitizer | |
| JP5570126B2 (en) | Method and apparatus for determining the posture of an object | |
| JP3624353B2 (en) | Three-dimensional shape measuring method and apparatus | |
| US8929608B2 (en) | Device and method for recognizing three-dimensional position and orientation of article | |
| US8121352B2 (en) | Fast three dimensional recovery method and apparatus | |
| US7986321B2 (en) | System and method for generating structured light for 3-dimensional image rendering | |
| TWI313576B (en) | Board inspecting apparatus, its parameter setting method and parameter setting apparatus | |
| CN105960569B (en) | The method of three-dimension object is checked using two dimensional image processing | |
| CN102713671A (en) | Point cloud data processing device, point cloud data processing method, and point cloud data processing program | |
| JP2012215394A (en) | Three-dimensional measuring apparatus and three-dimensional measuring method | |
| JP5342413B2 (en) | Image processing method | |
| EP3069100B1 (en) | 3d mapping device | |
| CN109976391B (en) | Three-dimensional space vision servo positioning system and method for focus photosensitive element of optical focusing lens | |
| CN104008366A (en) | 3D intelligent recognition method and system for biology | |
| EP1612509A1 (en) | Optical profilometer | |
| CN109341591A (en) | A kind of edge detection method and system based on handheld three-dimensional scanner | |
| US20140293009A1 (en) | Information processing apparatus and information processing method | |
| CN112037189A (en) | Device and method for detecting geometric parameters of steel bar welding seam | |
| US20130057652A1 (en) | Handheld Scanning Device | |
| Benveniste et al. | A color invariant for line stripe-based range scanners | |
| CN109751980A (en) | Wave height measurement method based on monocular vision laser triangulation | |
| CN118396944A (en) | A surface defect detection method and system based on computer vision | |
| CN115082538A (en) | 3D reconstruction system and method of multi-vision gimbal parts surface based on line structured light projection | |
| JP5336325B2 (en) | Image processing method | |
| Ozeki et al. | Real-time range measurement device for three-dimensional object recognition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140827 |
|
| RJ01 | Rejection of invention patent application after publication |