CN114972895A - Nondestructive detection method and device for crayfish quality - Google Patents
Nondestructive detection method and device for crayfish quality Download PDFInfo
- Publication number
- CN114972895A CN114972895A CN202210918448.6A CN202210918448A CN114972895A CN 114972895 A CN114972895 A CN 114972895A CN 202210918448 A CN202210918448 A CN 202210918448A CN 114972895 A CN114972895 A CN 114972895A
- Authority
- CN
- China
- Prior art keywords
- crayfish
- detected
- image
- model
- crayfishes
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A40/00—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production
- Y02A40/80—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in fisheries management
- Y02A40/81—Aquaculture, e.g. of fish
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Image Processing (AREA)
Abstract
The invention provides a nondestructive detection method and a nondestructive detection device for crayfish quality, and relates to the field of hyperspectral detection, wherein the method comprises the following steps: acquiring a three-dimensional color image of each wave band of the crayfish to be detected with the belly facing downwards; utilizing a Deep Net model to segment the region of interest to obtain a crayfish contour image, matching the crayfish contour image with a template library to determine a stylized crayfish image, and obtaining the target size of the crayfish to be detected; extracting spectral characteristics according to the three-dimensional color image, inputting the spectral characteristics into the trained convolutional neural network model, and outputting a prediction result of the content of the indexes to be detected of the crayfish to determine the maturity; screening the crayfishes with the weight, the volume and the maturity which reach the standard according to the target size and the maturity of the crayfishes to be detected; the template library comprises corresponding relations of each crayfish model, a plurality of back contour crayfish images with different angles and target sizes. The method solves the problems that weight classification is inaccurate, maturity cannot be accurately distinguished by naked eyes and the like, avoids the limitation of a conventional hyperspectral detection method, and has high detection accuracy.
Description
Technical Field
The invention relates to the field of hyperspectral detection, in particular to a nondestructive detection method and a nondestructive detection device for crayfish quality.
Background
In recent years, with the development of economy and the increase of consumption level, the quality requirement of crayfish for consumers is gradually increased. The high-quality crayfish is heavy, large in shrimp tail and good in maturity, and the quality of the crayfish is directly closely related to the use taste, the sale price and the like of the crayfish.
Currently, the classification of crayfish relies primarily on weighing to roughly sort it. Due to the physiological characteristics of the crayfish, the crayfish is easy to be stimulated to lose feet, clamp and the like in the process before the classification detection, some crayfish has a large head but a small tail, and some defective crayfish has large weight because the clamp is large, so that the crayfish cannot meet the requirement even if the weight meets the requirement and the quality cannot meet the requirement. In addition, the maturity of the crayfish mainly depends on manual work at present, the single maturity of the crayfish such as old and tender according to the hardness of the crayfish shell is judged, the requirements cannot be met, and overall, the quality of the crayfish has the problems of uneven classification, slow classification efficiency and inaccurate classification, and the requirements of enterprises and consumers are difficult to meet.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a nondestructive testing method and device for crayfish quality.
The invention provides a nondestructive testing method for crayfish quality, which comprises the following steps: acquiring two-bit gray level images of each wave band of the crayfish to be detected on the assembly line with the belly facing downwards through a spectral camera arranged at the top end of the assembly line, and synthesizing a three-dimensional color image according to the two-bit gray level images of all the wave bands;
identifying a crayfish body to be detected as an interested area based on the three-dimensional color image, segmenting the interested area by using a Deep Net model to obtain a contour image of the crayfish body, wherein the contour image comprises space angle information, comparing and matching the contour image with an established template library, determining a stylized crayfish image with the highest matching degree of contour and space angle in the template library, and determining the target size of the crayfish to be detected according to the actual crayfish target size corresponding to the stylized crayfish image;
according to the three-dimensional color image, extracting an original spectral curve by taking the whole crayfish body as an interested area, extracting an optimal characteristic wavelength combination as a spectral characteristic, inputting the spectral characteristic into a trained convolutional neural network model, outputting a crayfish index content prediction result to be detected, and determining the maturity of the crayfish according to the crayfish index content prediction result to be detected; the optimal characteristic wavelength is determined by extracting an original spectral curve according to a crayfish sample with known index content to be detected and screening the original spectral curve through the characteristic wavelength;
determining crayfishes meeting the standard conditions of weight and volume according to the target size of the crayfishes to be detected, determining the crayfishes meeting the standard conditions of maturity by combining the maturity, and screening the crayfishes meeting the standard conditions of weight, volume and maturity;
the template library comprises a plurality of crayfish models with different sizes, corresponding relations between each crayfish model and a plurality of stylized crayfish images with different space angle profiles, and corresponding relations between different crayfish models and the target size; the target size comprises an eye distance, an eye neck distance and a second section width of the tail; the index content to be detected comprises chitin content, astaxanthin content and protein content.
According to an embodiment of the invention, before acquiring a two-bit grayscale image of each wave band of the crayfish to be detected on the production line with the belly facing downwards by a spectral camera arranged at the top end of the production line, the method further comprises the following steps: establishing a three-dimensional geometric model of the actual crayfish, and obtaining a set of crayfish models with different sizes by adjusting the size of the model parameters; acquiring plane images of crayfishes with different sizes and with downward abdomen by adjusting different angles of each crayfish model, generating stylized crayfish images with corresponding sizes and angles, and determining the target size of each crayfish model; establishing a stylized crayfish image of each crayfish model and a plurality of different space angles with the abdomen facing down outline, establishing a corresponding relation between each stylized crayfish image and an actual target size obtained according to the crayfish model, marking the actual target size corresponding to each stylized image, and obtaining the established template library.
According to the nondestructive testing method for the crayfish quality, before the optimal characteristic wavelength combination is extracted as the spectral characteristic, the method further comprises the following steps: acquiring a plurality of three-dimensional color images of the crayfish sample with the determined index content to be detected, and dividing the three-dimensional color images into a training sample and a test sample respectively; screening the characteristic wavelength to obtain the screened characteristic wavelength of the training sample; inputting the spectral characteristics corresponding to the screened characteristic wavelengths of the training samples into the constructed convolutional neural network model for training; repeating the process of screening the characteristic wavelength until the constructed convolutional neural network model is input for a training process until the accuracy of the convolutional neural network model is highest after the characteristic wavelength verification training is finished after the test sample is screened; and taking the screened characteristic wavelength under the condition of highest accuracy as the optimal characteristic wavelength, and obtaining a corresponding convolutional neural network model as the trained convolutional neural network model.
According to the nondestructive detection method for the crayfish quality, the characteristic wavelength is screened, and the method comprises the following steps: screening initial characteristic wavelengths by using a competitive adaptive re-weighting sampling method CARS, and judging the influence degree of the initial characteristic wavelengths on the content of each index to be detected by using grey correlation; and selecting a plurality of characteristic wavelengths with the largest influence degree as the characteristic wavelengths after screening the content of each index to be detected.
According to the nondestructive detection method for crayfish quality, the method comprises the following steps of utilizing a Deep Net model to segment the region of interest to obtain a contour image of a crayfish body, wherein the contour image comprises space angle information, comparing and matching the contour image with an established template library, and before determining a stylized crayfish image with the highest matching degree of contour and space angle in the template library, further comprising: acquiring the three-dimensional color images of a plurality of crayfishes, extracting the region of interest of a crayfish body to obtain a corresponding crayfish body image, and marking the crayfish outline to obtain an outline training sample; inputting the outline training sample into the constructed initial Deep Net model, and training parameter updating of the initial Deep Net model according to the stylized crayfish image in the template library corresponding to the outline training sample and the model output result; and repeating the process of training the initial Deep Net model until a preset standard reaching condition is met, so as to obtain the Deep Net model.
According to the nondestructive detection method for the quality of the crayfish, the crayfish meeting the standard conditions of weight and volume is determined according to the target size of the crayfish to be detected, and the nondestructive detection method comprises the following steps: according to the three conditions that the eye distance is larger than 1.4mm, the eye neck distance is larger than 2.3 mm and the second pitch width of the tail is larger than 0.9mm, the conditions of the eye distance and the second pitch width of the tail are met simultaneously, or the conditions of the eye neck distance and the second pitch width of the tail are met simultaneously, or the conditions of the eye distance, the eye neck distance and the second pitch width of the tail are met simultaneously, the crayfishes meeting the weight and volume standard reaching conditions are determined.
The invention also provides a nondestructive testing device for the quality of crayfishes, which comprises: the image acquisition module is used for acquiring two-bit gray level images of each wave band of the crayfish to be detected on the assembly line with the belly facing downwards through a spectral camera arranged at the top end of the assembly line, and synthesizing a three-dimensional color image according to the two-bit gray level images of all the wave bands;
the model matching module is used for identifying the crayfish body to be detected as an interested area based on the three-dimensional color image, utilizing a Deep Net model to divide the interested area to obtain a contour image of the crayfish body, wherein the contour image comprises space angle information, comparing and matching the contour image with an established template library, determining a stylized crayfish image with the highest matching degree of contour and space angle in the template library, and determining the target size of the crayfish to be detected according to the actual crayfish target size corresponding to the stylized crayfish image;
the spectrum processing module is used for extracting an original spectrum curve by taking the whole crawfish body as an interested region according to the three-dimensional color image, extracting an optimal characteristic wavelength combination as a spectrum characteristic, inputting the spectrum characteristic into the trained convolutional neural network model, outputting a prediction result of the content of the crawfish to be detected indexes, and determining the maturity of the crawfish according to the prediction result of the content of the crawfish to be detected indexes; the optimal characteristic wavelength is determined by extracting an original spectral curve according to a crayfish sample with known index content to be detected and screening the original spectral curve through the characteristic wavelength;
the comprehensive screening module is used for determining the crayfishes meeting the standard conditions of weight and volume according to the target size of the crayfishes to be detected, determining the crayfishes meeting the standard conditions of maturity by combining the maturity, and screening the crayfishes meeting the standards of weight, volume and maturity;
the template library comprises a plurality of crayfish models with different sizes, corresponding relations between each crayfish model and stylized crayfish images of back outlines with different space angles, and corresponding relations between different crayfish models and the target size; the target size comprises an eye distance, an eye neck distance and a second section width of the tail; the index content to be detected comprises chitin content, astaxanthin content and protein content.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the program to realize the nondestructive detection method for the quality of the crayfish.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method for nondestructive testing of crayfish quality as described in any of the above.
The invention also provides a nondestructive testing system for the quality of crayfish, comprising: the device comprises an attitude adjusting mechanism, an annular adjustable light source, a hyperspectral imager, a camera bellows, a PLC (programmable logic controller) control unit, a transmission belt and the electronic equipment; the posture adjusting mechanism is used for adjusting the spatial position and angle of the crayfish to be detected so that the belly of the crayfish to be detected faces downwards; the hyperspectral imager is arranged in the camera bellows, is connected with the electronic equipment through a data line and is used for shooting hyperspectral images when the crayfishes to be detected pass through the camera bellows; the annular adjustable light source provides a light source for acquiring the hyperspectral image; the conveyor belt is used for conveying the crayfishes to be detected; and the PLC control unit is used for controlling the movement of the conveying belt.
The nondestructive detection method and the nondestructive detection device for the quality of the crayfishes solve the problems that weight grading is inaccurate and maturity can not be accurately distinguished by naked eyes due to the fact that the crayfishes lose feet due to external stimulation and the like, high-quality crayfishes which are large in weight, large in size and good in maturity and meet actual requirements are obtained by combining the highlight three-dimensional color image through the corresponding relation of the outline image, the stylized crayfishes image and the target size, and limitation of a conventional hyperspectral detection method is avoided, so that the detection accuracy is high, and the detection result is more objective and reliable. The method can realize the classified prediction of the high-quality crayfishes without destroying the crayfishes and judging by naked eyes, realizes the classified prediction of the high-quality crayfishes, has high detection efficiency and low detection cost, and is favorable for the automatic treatment of a production line.
Drawings
In order to more clearly illustrate the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a schematic flow chart of a nondestructive testing method for crayfish quality provided by the present invention;
FIG. 2 is a second schematic flow chart of the nondestructive testing method for crayfish quality provided by the present invention;
FIG. 3 is a schematic structural diagram of a nondestructive testing device for crayfish quality provided by the invention;
FIG. 4 is a schematic structural diagram of an electronic device provided by the present invention;
FIG. 5 is a schematic structural diagram of a nondestructive testing system for crayfish quality provided by the invention;
description of reference numerals: 1. the crayfish to be detected; 2. an attitude adjusting mechanism; 3. an annular tunable light source; 4. a hyperspectral imager; 5. a dark box; 6. an electronic device; 7. a PLC control unit; 8. and (5) conveying the belt.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The nondestructive testing method and device for crayfish quality of the invention are described below with reference to fig. 1 to 5. Fig. 1 is a schematic flow diagram of a nondestructive testing method for crayfish quality, as shown in fig. 1, the nondestructive testing method for crayfish quality includes:
101. and acquiring a two-bit gray image of each wave band of the crayfish to be detected on the assembly line with the belly facing downwards through a spectral camera arranged at the top end of the assembly line, and synthesizing a three-dimensional color image according to the two-bit gray images of all the wave bands.
The detection sample can be crayfish with any weight, volume and maturity, and various crayfish can be randomly detected at the same time.
The crayfish is operated below the hyperspectral detection device after being subjected to posture adjustment on a transmission line of the assembly line, and the posture adjustment mainly adjusts the posture to be that the belly faces downwards, so that spectrum back images can be collected conveniently. And adjusting parameters of the detection device, including the focal length, the exposure time and the like of the push-broom hyperspectral imager lens with the standard lens. And determining a corresponding two-dimensional gray-scale image under each wave band. And synthesizing the two-dimensional gray level images of all the wave bands into a three-dimensional color hyperspectral image.
102. Based on the three-dimensional color image, recognizing the crayfish body to be detected as an interested area, utilizing a Deep Net model to segment the interested area to obtain a contour image of the crayfish body, wherein the contour image comprises space angle information, comparing and matching the contour image with an established template library, determining a stylized crayfish image with the highest matching degree of contour and space angle in the template library, and determining the target size of the crayfish to be detected according to the actual crayfish target size corresponding to the stylized crayfish image.
The whole crayfish body is used as an interested area to identify the crayfish body, and the divided crayfish outline image is automatically compared and matched with the stylized crayfish image in the template library by using the Deep Net model in the embodiment of the invention. The Deep Net model is trained before use, such as training with an image of the body of a crayfish labeled with a contour.
The template library comprises a plurality of crayfish models with different sizes, corresponding relations between each crayfish model and a plurality of stylized crayfish images with different space angle profiles, and corresponding relations between different crayfish models and the target size; the target size comprises an eye distance, an eye neck distance and a second section width of the tail; the index content to be detected comprises chitin content, astaxanthin content and protein content.
In the invention, the template library establishes the corresponding relation between each crawfish model with different sizes and a plurality of back contour images with different angles or postures in advance, and generates stylized images and the corresponding relation between the different stylized crawfish images and the target size.
Firstly, determining a plurality of crayfishes with different sizes or shapes, and then determining specific values of target dimensions such as the eye distance, the eye-neck distance, the second section width of the tail and the like for each crayfish to form a one-to-one correspondence relationship. Then, for each crawfish of different size or shape, a crawfish back contour image of different angle or pose is also determined. The crayfish model with different angles or postures can be adjusted through each crayfish model with different sizes or shapes, then the complete contour image with the belly of the crayfish facing downwards with different angles or postures is extracted according to the model, and therefore each crayfish with different sizes or shapes corresponds to the complete contour images with the belly of the crayfish facing downwards with different angles or postures, and a one-to-many relationship is formed.
In the one-to-many relationship, the changing step length of the crayfishes with different angles or postures can be set according to the requirement. Usually within a limited step size, the eye distance, the eye neck distance, the width of the second section of the tail and the like do not change much.
And selecting a matched candidate stylized crayfish image from the template library according to the contour image determined by the step 102. Specifically, a crayfish abdomen downward contour image matched with a certain angle posture is determined according to the contour image, so that a candidate stylized crayfish image corresponding to the crayfish abdomen downward contour image is matched, the target size of a matched candidate crayfish model is predetermined and marked, the target size of the crayfish corresponding to the contour image in 102 is obtained, namely the target sizes such as the accurate eye distance, the eye neck distance and the second pitch width of the tail are determined according to the contour image, and the accurate weight and the size can be determined in 104 according to the accurate target size.
103. According to the three-dimensional color image, extracting an original spectral curve by taking the whole crayfish body as an interested area, extracting an optimal characteristic wavelength combination as a spectral characteristic, inputting the spectral characteristic into a trained convolutional neural network model, outputting a crayfish index content prediction result to be detected, and determining the maturity of the crayfish according to the crayfish index content prediction result to be detected; and the optimal characteristic wavelength is determined by extracting an original spectrum curve according to a crayfish sample with known index content to be detected and screening the original spectrum curve through the characteristic wavelength.
In 103, the original spectral curve is a corresponding curve of each wavelength and reflectivity in the three-dimensional color image. The characteristic wavelength is a part of wavelengths in the original spectrum curve, and the spectrum characteristic is a reflectivity value of the corresponding wavelength. For the convolutional neural network model, after the spectral characteristics are extracted in advance according to the crayfish sample with known content of the index to be detected according to the method, the spectral characteristics are used as input, and a large amount of training is carried out by using the known content of the index to be detected as a label, so that the trained convolutional neural network model is obtained.
And inputting the spectral characteristics extracted in the step 103 into the trained convolutional neural network model to obtain an accurate prediction result of the content of the index to be detected, wherein the content of the index to be detected comprises chitin content, astaxanthin content, meat content and the like.
104. Determining the crayfishes meeting the standard conditions of weight and volume according to the target size of the crayfishes to be detected, determining the crayfishes meeting the standard conditions of maturity by combining the maturity, and screening the crayfishes meeting the standard conditions of weight, volume and maturity.
According to the condition of the target size and the maturity, the crayfishes with large volume, heavy weight and good maturity are determined.
In one embodiment, the determining the crayfish meeting the weight and volume standard reaching conditions according to the target size of the crayfish to be detected comprises the following steps: according to the three conditions that the eye distance is larger than 1.4mm, the eye neck distance is larger than 2.3 mm and the second width of the tail is larger than 0.9mm, the conditions of the eye distance and the second width of the tail are met at the same time, or the conditions of the eye distance, the eye neck distance and the second width of the tail are met at the same time, the crayfish meeting the weight and volume standard reaching conditions is determined.
For eye distances greater than 1.4mm, eye neck distances greater than 2.3 mm and tail second pitch widths greater than 0.9 mm. The preset condition can be that the eye distance and the second section width of the tail are met simultaneously, or the eye neck distance and the second section width of the tail are met simultaneously, or the three are met simultaneously, so that the crayfish with heavy weight and large volume can be determined. In addition, maturity conditions meeting the contents of chitin, astaxanthin and the like at the same time need to be determined, and based on the maturity conditions, the high-quality crayfish with large volume, heavy weight and good maturity is obtained.
The nondestructive detection method for the quality of the crayfish solves the problems that weight grading is inaccurate and maturity can not be accurately distinguished by naked eyes due to the fact that the crayfish loses feet due to external stimulation and the like, high-quality crayfish which is large in weight, large in size and good in maturity and meets the actual requirements is obtained through the corresponding relation of the outline image, the stylized crayfish image and the target size and the highlight three-dimensional color image, and the limitation of a conventional hyperspectral detection method is avoided, so that the detection accuracy is high, and the detection result is objective and reliable. The method can realize the classified prediction of the high-quality crayfishes without destroying the crayfishes and judging by naked eyes, realizes the classified prediction of the high-quality crayfishes, has high detection efficiency and low detection cost, and is favorable for the automatic treatment of a production line.
In one embodiment, before acquiring the two-bit grayscale image of each wave band of the crayfish to be detected on the pipeline with the abdomen facing downwards through the spectral camera arranged at the top end of the pipeline, the method further comprises the following steps: establishing a three-dimensional geometric model of the actual crayfish, and obtaining a set of crayfish models with different sizes by adjusting the size of the model parameters; acquiring plane images of crayfishes of different sizes with the abdomen facing downwards by adjusting different angles of each crayfish model, generating stylized crayfish images of corresponding sizes and angles, and determining the target size of each crayfish model; establishing stylized crayfish images of each crayfish model and a plurality of abdomen downward outlines with different space angles, corresponding relations between the stylized crayfish images and actual target sizes obtained according to the crayfish models, marking the actual target sizes corresponding to the stylized images, and obtaining the established template library.
Specifically, a three-dimensional software can be utilized to establish a solid geometric model highly reduced with the actual crayfish appearance. And then, obtaining the geometrical models of the crayfishes with different sizes through parameter adjustment. And extracting the eye distance, the eye neck distance and the second section width of the tail of the corresponding crayfish in each model, wherein the eye distance, the eye neck distance and the second section width of the tail are approximately equal due to the high reducibility of the shapes. And rotating each crayfish geometric model to obtain back plane images of crayfishes with different sizes and with the belly facing downwards, generating stylized crayfish images of corresponding sizes, angles and postures, and extracting the contours to obtain contour images of different spatial angles. Marking the actual crayfish eye distance, eye neck distance and second tail section width corresponding to each model, or marking the actual crayfish eye distance, eye neck distance and second tail section width corresponding to the back contour images in different angles or postures to obtain the template library.
According to the nondestructive testing method for the crayfish quality, provided by the embodiment of the invention, the crayfish models with different sizes are obtained by adjusting the model parameters, modeling of the crayfish models is not needed, and the modeling workload is reduced; through adjusting different angles or gestures, a plurality of abdomen downward profile images are obtained, the problem that each crayfish is difficult to adjust to the standard gesture on the assembly line can be adapted, the crayfish with different angles and gestures can be matched with an accurate crayfish model, the detection precision of the target size is further improved, and the weight and the volume of the crayfish with high precision are obtained. In addition, through the constructed template library, the corresponding target size can be accurately matched only through the contour image, the calculation amount overhead of image processing is greatly reduced, high-speed detection on a production line is facilitated, and the effect of high detection precision is realized.
In one embodiment, before extracting the optimal characteristic wavelength combination as the spectral characteristic, the method further includes: acquiring a plurality of three-dimensional color images of the crayfish sample with the determined index content to be detected, and dividing the three-dimensional color images into a training sample and a test sample respectively; screening the characteristic wavelength to obtain the screened characteristic wavelength of the training sample; inputting the spectral characteristics corresponding to the screened characteristic wavelengths of the training samples into the constructed convolutional neural network model for training; repeating the process of screening the characteristic wavelength until the constructed convolutional neural network model is input for a training process until the accuracy of the convolutional neural network model is highest after the characteristic wavelength verification training is finished after the test sample is screened; and taking the screened characteristic wavelength under the condition of highest accuracy as the optimal characteristic wavelength, and obtaining a corresponding convolutional neural network model as the trained convolutional neural network model.
Specifically, in the embodiment of the invention, the characteristic wavelength is continuously screened in the training process, the detection precision of the convolutional neural network model is verified, and the corresponding characteristic wavelength and the detection model are determined under the condition of the highest precision, so that the high-accuracy detection of the content of the chitin and the content of the astaxanthin is realized.
In one embodiment, the screening the characteristic wavelength includes: screening initial characteristic wavelengths by using a competitive adaptive re-weighting sampling method CARS, and judging the influence degree of the initial characteristic wavelengths on the content of each index to be detected by using grey correlation; and selecting a plurality of characteristic wavelengths with the largest influence degree as the characteristic wavelengths after screening the content of each index to be detected.
Fig. 2 is a second schematic flow chart of the nondestructive testing method for crayfish quality provided by the present invention, and as shown in fig. 2, in the specific process of feature screening, the present invention first screens the initial feature wavelength by CARS. For example, a PLS model of the relationship of wavelengths to chitin and astaxanthin can be established from a plurality of training samples of known chitin and astaxanthin. And (2) reserving a point with larger regression coefficient absolute value weight in the PLS model as a new subset through adaptive weighted sampling (ARS) each time, removing the point with smaller weight, then establishing the PLS model based on the new subset, and selecting the wavelength in the subset with the minimum PLS model interactive verification Root Mean Square Error (RMSECV) as the initial characteristic wavelength through multiple calculations.
On the basis, the initial characteristic wavelength is continuously screened, and the influence degree of the initial characteristic wavelength on the chitin and the astaxanthin is judged according to the grey correlation degree. The gray correlation method is a method for measuring the degree of correlation between factors according to the similarity or difference of the development trends between the factors, i.e., "gray correlation degree". When the index value to be detected is the astaxanthin, the gray relevance degree is utilized to screen out the characteristic wavelength which is strongly related to the chitin, the characteristic wavelength is used as the spectral characteristic to be input into the convolutional neural network model, when the index value to be detected is the astaxanthin, the gray relevance degree is utilized to screen out the characteristic wavelength which is strongly related to the astaxanthin, the characteristic wavelength is used as the spectral characteristic to be input into the convolutional neural network model, namely, the model can automatically select the optimal wavelength which is closely related to the index to be detected as the characteristic wavelength according to the difference of the index to be detected, and therefore the prediction capability of the model can be well improved.
In an embodiment, the segmenting the region of interest by using the Deep Net model to obtain a contour image of the crawfish body, the contour image including spatial angle information, and comparing and matching the contour image with an established template library, and before determining a stylized crawfish image with the highest matching degree between the contour and the spatial angle in the template library, the method further includes: acquiring the three-dimensional color images of a plurality of crayfishes, extracting the region of interest of a crayfish body to obtain a corresponding crayfish body image, and marking the crayfish outline to obtain an outline training sample; inputting the outline training sample into the constructed initial Deep Net model, and training parameter updating of the initial Deep Net model according to the stylized crayfish image in the template library corresponding to the outline training sample and the model output result; and repeating the process of training the initial Deep Net model until a preset standard reaching condition is met, so as to obtain the Deep Net model.
The three-dimensional color images of the crayfishes are obtained by the same method, the region of interest of the crayfish body is extracted, the corresponding crayfish body image is obtained, the crayfish contour is marked, the contour training sample is obtained, then the built Deep Net model is trained, the specific training process can refer to the prior art, and the detailed description is omitted here.
The nondestructive testing device for crayfish quality provided by the present invention is described below, and the nondestructive testing device for crayfish quality described below and the nondestructive testing method for crayfish quality described above can be referred to in correspondence with each other.
Fig. 3 is a schematic structural diagram of the nondestructive testing apparatus for crayfish quality provided by the present invention, and as shown in fig. 3, the nondestructive testing apparatus for crayfish quality includes: an image acquisition module 301, a model matching module 302, a spectral processing module 303, and a comprehensive screening module 304. The image acquisition module 301 is used as an image acquisition module and is used for acquiring a two-bit gray level image of each wave band of the crayfish to be detected on the assembly line with the belly facing downwards through a spectral camera arranged at the top end of the assembly line and synthesizing a three-dimensional color image according to the two-bit gray level images of all the wave bands; the model matching module 302 is used for identifying a crayfish body to be detected as an interested area based on the three-dimensional color image, utilizing a Deep Net model to segment the interested area to obtain a contour image of the crayfish body, wherein the contour image comprises space angle information, comparing and matching the contour image with an established template library, determining a stylized crayfish image with the highest matching degree of the contour and the space angle in the template library, and determining the target size of the crayfish to be detected according to the actual crayfish target size corresponding to the stylized crayfish image; the spectrum processing module 303 is used for extracting an original spectrum curve by taking the whole crawfish body as an interested region according to the three-dimensional color image, extracting an optimal characteristic wavelength combination as a spectrum characteristic, inputting the spectrum characteristic into a trained convolutional neural network model, outputting a prediction result of the content of the crawfish to be detected, and determining the maturity of the crawfish according to the prediction result of the content of the crawfish to be detected; the optimal characteristic wavelength is determined by extracting an original spectral curve according to a crayfish sample with known index content to be detected and screening the original spectral curve through the characteristic wavelength; the comprehensive screening module 304 is used for determining the crayfishes meeting the standard conditions of weight and volume according to the target size of the crayfishes to be detected, and determining the crayfishes meeting the standard conditions of maturity by combining the maturity so as to screen the crayfishes meeting the standard conditions of weight, volume and maturity.
The device embodiment provided in the embodiments of the present invention is for implementing the above method embodiments, and for details of the process and the details, reference is made to the above method embodiments, which are not described herein again.
The implementation principle and the generated technical effect of the nondestructive testing device for the quality of the crayfish provided by the embodiment of the invention are the same as those of the nondestructive testing method for the quality of the crayfish, and for brief description, the corresponding contents in the nondestructive testing method for the quality of the crayfish can be referred to for the parts which are not mentioned in the embodiment of the nondestructive testing device for the quality of the crayfish.
Fig. 4 is a schematic structural diagram of an electronic device provided in the present invention, and as shown in fig. 4, the electronic device may include: a processor (processor)401, a communication Interface (communication Interface)402, a memory (memory)403 and a communication bus 404, wherein the processor 401, the communication Interface 402 and the memory 403 complete communication with each other through the communication bus 404. Processor 401 may invoke logic instructions in memory 403 to perform a method for nondestructive testing of crawfish quality comprising: acquiring two-bit gray level images of each wave band of the crayfish to be detected on the assembly line with the belly facing downwards through a spectral camera arranged at the top end of the assembly line, and synthesizing a three-dimensional color image according to the two-bit gray level images of all the wave bands;
identifying a crayfish body to be detected as an interested area based on the three-dimensional color image, segmenting the interested area by using a Deep Net model to obtain a contour image of the crayfish body, wherein the contour image comprises space angle information, comparing and matching the contour image with an established template library, determining a stylized crayfish image with the highest matching degree of contour and space angle in the template library, and determining the target size of the crayfish to be detected according to the actual crayfish target size corresponding to the stylized crayfish image;
according to the three-dimensional color image, extracting an original spectral curve by taking the whole crayfish body as an interested area, extracting an optimal characteristic wavelength combination as a spectral characteristic, inputting the spectral characteristic into a trained convolutional neural network model, outputting a crayfish index content prediction result to be detected, and determining the maturity of the crayfish according to the crayfish index content prediction result to be detected; the optimal characteristic wavelength is determined by extracting an original spectral curve according to a crayfish sample with known index content to be detected and screening the original spectral curve through the characteristic wavelength;
determining crayfishes meeting the standard conditions of weight and volume according to the target size of the crayfishes to be detected, determining the crayfishes meeting the standard conditions of maturity by combining the maturity, and screening the crayfishes meeting the standard conditions of weight, volume and maturity;
the template library comprises a plurality of crayfish models with different sizes, corresponding relations between each crayfish model and a plurality of stylized crayfish images with different space angle profiles, and corresponding relations between different crayfish models and the target size; the target size comprises an eye distance, an eye neck distance and a second section width of the tail; the index content to be detected comprises chitin content, astaxanthin content and protein content.
In addition, the logic instructions in the memory 403 may be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention or a part thereof which substantially contributes to the prior art may be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In another aspect, the present invention further provides a non-transitory computer readable storage medium, on which a computer program is stored, the computer program being implemented by a processor to execute the nondestructive testing method for crawfish quality provided by the above embodiments, the method comprising: acquiring two-bit gray level images of each wave band of the crayfishes to be detected on the assembly line with the belly downward through a spectral camera arranged at the top end of the assembly line, and synthesizing a three-dimensional color image according to the two-bit gray level images of all the wave bands;
identifying a crayfish body to be detected as an interested area based on the three-dimensional color image, segmenting the interested area by using a Deep Net model to obtain a contour image of the crayfish body, wherein the contour image comprises space angle information, comparing and matching the contour image with an established template library, determining a stylized crayfish image with the highest matching degree of contour and space angle in the template library, and determining the target size of the crayfish to be detected according to the actual crayfish target size corresponding to the stylized crayfish image;
according to the three-dimensional color image, extracting an original spectral curve by taking the whole crayfish body as an interested area, extracting an optimal characteristic wavelength combination as a spectral characteristic, inputting the spectral characteristic into a trained convolutional neural network model, outputting a crayfish index content prediction result to be detected, and determining the maturity of the crayfish according to the crayfish index content prediction result to be detected; the optimal characteristic wavelength is determined by extracting an original spectral curve according to a crayfish sample with known index content to be detected and screening the original spectral curve through the characteristic wavelength;
determining crayfishes meeting the standard conditions of weight and volume according to the target size of the crayfishes to be detected, determining the crayfishes meeting the standard conditions of maturity by combining the maturity, and screening the crayfishes meeting the standard conditions of weight, volume and maturity;
the template library comprises a plurality of crayfish models with different sizes, corresponding relations between each crayfish model and a plurality of stylized crayfish images with different space angle profiles, and corresponding relations between different crayfish models and the target size; the target size comprises an eye distance, an eye neck distance and a second section width of the tail; the index content to be detected comprises chitin content, astaxanthin content and protein content.
Fig. 5 is a schematic structural diagram of the nondestructive testing system for crayfish quality provided by the present invention, and as shown in fig. 5, the nondestructive testing system for crayfish quality includes: the device comprises an attitude adjusting mechanism 2, an annular adjustable light source 3, a hyperspectral imager 4, a camera bellows 5, the electronic equipment 6 of the embodiment, a PLC control unit 7 and a transmission belt 8. The posture adjusting mechanism 2 is used for adjusting the spatial position and the angle of the crayfish 1 to be detected so that the belly of the crayfish 1 to be detected faces downwards; the hyperspectral imager 4 is arranged in the camera bellows 5, and the hyperspectral imager 4 is connected with the electronic equipment 6 through a data line and is used for shooting a hyperspectral image when the crayfish to be detected passes through the camera bellows 5; the annular adjustable light source 3 is used for providing a light source for acquiring the hyperspectral image; the conveyor belt 8 is used for conveying the crayfish 1 to be detected; the PLC control unit 7 is used for controlling the movement of the conveying belt.
The hyperspectral imager 4 can be connected with the electronic equipment 6 through a USB3.0 data line, the push-broom hyperspectral imager 4 with a standard lens adopts an onboard hyperspectral camera, the model can be OCI-UAV-1000, the data acquisition mode is a push-broom type, and the size of the camera lens (35 mm fixed focus, 18-degree field angle) is 80 mm x 60 mm x 60 mm; its function is to take a high quality image of crayfish.
The annular adjustable light source 3 is mounted in a dark box 5, which may be of the type MV-LBES-300-W, Haekwover, with a power of 65W. The electric energy of the annular adjustable light source can be obtained from the electronic equipment 6 through the data line, and the light source is provided for obtaining the hyperspectral image.
The PLC control unit 7 is a universal component and is used for controlling the movement of the conveying belt. Its function of transmission band 8 is mainly the transmission of the cray of waiting to examine for the cray can reach each assigned position and detect. For specific processing procedures and technical effects, reference may be made to the above method embodiments, which are not described herein again.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment may be implemented by software plus a necessary general hardware platform, and may also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A nondestructive detection method for crayfish quality is characterized by comprising the following steps:
acquiring two-bit gray level images of each wave band of the crayfish to be detected on the assembly line with the belly facing downwards through a spectral camera arranged at the top end of the assembly line, and synthesizing a three-dimensional color image according to the two-bit gray level images of all the wave bands;
identifying a crayfish body to be detected as an interested area based on the three-dimensional color image, segmenting the interested area by using a Deep Net model to obtain a contour image of the crayfish body, wherein the contour image comprises space angle information, comparing and matching the contour image with an established template library, determining a stylized crayfish image with the highest matching degree of contour and space angle in the template library, and determining the target size of the crayfish to be detected according to the actual crayfish target size corresponding to the stylized crayfish image;
according to the three-dimensional color image, extracting an original spectral curve by taking the whole crayfish body as an interested area, extracting an optimal characteristic wavelength combination as a spectral characteristic, inputting the spectral characteristic into a trained convolutional neural network model, outputting a crayfish index content prediction result to be detected, and determining the maturity of the crayfish according to the crayfish index content prediction result to be detected; the optimal characteristic wavelength is determined by extracting an original spectral curve according to a crayfish sample with known index content to be detected and screening the original spectral curve through the characteristic wavelength;
determining crayfishes meeting the standard conditions of weight and volume according to the target size of the crayfishes to be detected, determining the crayfishes meeting the standard conditions of maturity by combining the maturity, and screening the crayfishes meeting the standard conditions of weight, volume and maturity;
the template library comprises a plurality of crayfish models with different sizes, corresponding relations between each crayfish model and a plurality of stylized crayfish images with different space angle profiles, and corresponding relations between different crayfish models and the target size; the target size comprises an eye distance, an eye neck distance and a second section width of the tail; the index content to be detected comprises chitin content, astaxanthin content and protein content.
2. The nondestructive testing method for the quality of the crayfish as claimed in claim 1, wherein before the acquiring of the two-bit gray scale image of each wave band of the crayfish to be tested with the belly facing downwards on the assembly line by the spectral camera arranged at the top end of the assembly line, the nondestructive testing method further comprises:
establishing a three-dimensional geometric model of the actual crayfish, and obtaining a set of crayfish models with different sizes by adjusting the size of the model parameters;
acquiring plane images of crayfishes with different sizes and with downward abdomen by adjusting different angles of each crayfish model, generating stylized crayfish images with corresponding sizes and angles, and determining the target size of each crayfish model;
establishing a stylized crayfish image of each crayfish model and a plurality of different space angles with the abdomen facing down outline, establishing a corresponding relation between each stylized crayfish image and an actual target size obtained according to the crayfish model, marking the actual target size corresponding to each stylized image, and obtaining the established template library.
3. The method for nondestructive testing of crayfish quality as claimed in claim 1, wherein before said extracting the optimal characteristic wavelength combination as the spectral characteristic, the method further comprises:
acquiring a plurality of three-dimensional color images of the crayfish sample with the determined index content to be detected, and dividing the three-dimensional color images into a training sample and a test sample respectively;
screening the characteristic wavelength to obtain the screened characteristic wavelength of the training sample;
inputting the spectral characteristics corresponding to the screened characteristic wavelengths of the training samples into the constructed convolutional neural network model for training;
repeating the process of screening the characteristic wavelength until the constructed convolutional neural network model is input for a training process until the accuracy of the convolutional neural network model is highest after the characteristic wavelength verification training is finished after the test sample is screened;
and taking the screened characteristic wavelength under the condition of highest accuracy as the optimal characteristic wavelength, and obtaining a corresponding convolutional neural network model as the trained convolutional neural network model.
4. The nondestructive testing method for the quality of crayfish according to claim 3, wherein said screening for the characteristic wavelength includes:
screening initial characteristic wavelengths by using a competitive adaptive re-weighting sampling method CARS, and judging the influence degree of the initial characteristic wavelengths on the content of each index to be detected by using grey correlation;
and selecting a plurality of characteristic wavelengths with the largest influence degree as the characteristic wavelengths after screening the content of each index to be detected.
5. The nondestructive testing method for the quality of the crayfish as claimed in claim 1, wherein the step of segmenting the region of interest by using a Deep Net model to obtain a contour image of the crayfish body including spatial angle information, comparing and matching the contour image with an established template library, and before determining the stylized crayfish image with the highest matching degree between the contour and the spatial angle in the template library further comprises the steps of:
acquiring the three-dimensional color images of a plurality of crayfishes, extracting the region of interest of a crayfish body to obtain a corresponding crayfish body image, and marking the crayfish outline to obtain an outline training sample;
inputting the outline training sample into the constructed initial Deep Net model, and training parameter updating of the initial Deep Net model according to the stylized crayfish image in the template library corresponding to the outline training sample and the model output result;
and repeating the process of training the initial Deep Net model until a preset standard reaching condition is met, so as to obtain the Deep Net model.
6. The nondestructive testing method for the quality of crayfish as claimed in claim 1, wherein the step of determining the crayfish satisfying the weight and volume standard conditions according to the target size of the crayfish to be tested comprises the following steps:
according to the three conditions that the eye distance is larger than 1.4mm, the eye neck distance is larger than 2.3 mm and the second pitch width of the tail is larger than 0.9mm, the conditions of the eye distance and the second pitch width of the tail are met simultaneously, or the conditions of the eye neck distance and the second pitch width of the tail are met simultaneously, or the conditions of the eye distance, the eye neck distance and the second pitch width of the tail are met simultaneously, the crayfishes meeting the weight and volume standard reaching conditions are determined.
7. The utility model provides a crayfish quality nondestructive test device which characterized in that includes:
the image acquisition module is used for acquiring two-bit gray level images of each wave band of the crayfish to be detected on the assembly line with the belly facing downwards through a spectral camera arranged at the top end of the assembly line, and synthesizing a three-dimensional color image according to the two-bit gray level images of all the wave bands;
the model matching module is used for identifying the crayfish body to be detected as an interested area based on the three-dimensional color image, utilizing a Deep Net model to divide the interested area to obtain a contour image of the crayfish body, wherein the contour image comprises space angle information, comparing and matching the contour image with an established template library, determining a stylized crayfish image with the highest matching degree of contour and space angle in the template library, and determining the target size of the crayfish to be detected according to the actual crayfish target size corresponding to the stylized crayfish image;
the spectrum processing module is used for extracting an original spectrum curve by taking the whole crawfish body as an interested region according to the three-dimensional color image, extracting an optimal characteristic wavelength combination as a spectrum characteristic, inputting the spectrum characteristic into the trained convolutional neural network model, outputting a prediction result of the content of the crawfish to be detected indexes, and determining the maturity of the crawfish according to the prediction result of the content of the crawfish to be detected indexes; the optimal characteristic wavelength is determined by extracting an original spectral curve according to a crayfish sample with known index content to be detected and screening the original spectral curve through the characteristic wavelength;
the comprehensive screening module is used for determining the crayfishes meeting the standard conditions of weight and volume according to the target size of the crayfishes to be detected, determining the crayfishes meeting the standard conditions of maturity by combining the maturity, and screening the crayfishes meeting the standards of weight, volume and maturity;
the template library comprises a plurality of crayfish models with different sizes, corresponding relations between each crayfish model and stylized crayfish images of back profiles with different space angles, and corresponding relations between different crayfish models and the target size; the target size comprises an eye distance, an eye neck distance and a second section width of the tail; the index content to be detected comprises chitin content, astaxanthin content and protein content.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of nondestructive testing of crawfish quality as claimed in any of claims 1 to 6 when executing the program.
9. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the method for nondestructive testing of crawfish quality as claimed in any one of claims 1 to 6.
10. A crawfish quality nondestructive testing system, comprising:
an attitude adjustment mechanism, an annular adjustable light source, a hyperspectral imager, a camera bellows, a PLC control unit, a conveyor belt and the electronic device of claim 8;
the posture adjusting mechanism is used for adjusting the spatial position and the angle of the crayfish to be detected so that the belly of the crayfish to be detected faces downwards;
the hyperspectral imager is arranged in the camera bellows, is connected with the electronic equipment through a data line and is used for shooting hyperspectral images when the crayfishes to be detected pass through the camera bellows;
the annular adjustable light source provides a light source for acquiring the hyperspectral image;
the conveyor belt is used for conveying the crayfishes to be detected;
and the PLC control unit is used for controlling the movement of the conveying belt.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210918448.6A CN114972895B (en) | 2022-08-01 | 2022-08-01 | Nondestructive testing method and device for crayfish quality |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210918448.6A CN114972895B (en) | 2022-08-01 | 2022-08-01 | Nondestructive testing method and device for crayfish quality |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114972895A true CN114972895A (en) | 2022-08-30 |
| CN114972895B CN114972895B (en) | 2022-09-30 |
Family
ID=82969054
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210918448.6A Active CN114972895B (en) | 2022-08-01 | 2022-08-01 | Nondestructive testing method and device for crayfish quality |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114972895B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117299596A (en) * | 2023-08-14 | 2023-12-29 | 江苏秦郡机械科技有限公司 | Material screening system and method for automatic detection |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111832532A (en) * | 2020-07-24 | 2020-10-27 | 上海电气集团自动化工程有限公司 | Online visual detection method and system for crayfish pose identification |
| WO2021158894A1 (en) * | 2020-02-07 | 2021-08-12 | Nova-Tech Engineering, Llc | Shrimp processing system, processing apparatus and methods |
| US20210321593A1 (en) * | 2020-04-21 | 2021-10-21 | InnovaSea Systems, Inc. | Systems and methods for fish volume estimation, weight estimation, and analytic value generation |
| CN113537175A (en) * | 2021-09-16 | 2021-10-22 | 长沙鹏阳信息技术有限公司 | Same-fence swinery average weight estimation method based on computer vision |
| CN114399674A (en) * | 2022-01-21 | 2022-04-26 | 岭南师范学院 | Hyperspectral image technology-based shellfish toxin nondestructive rapid detection method and system |
-
2022
- 2022-08-01 CN CN202210918448.6A patent/CN114972895B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021158894A1 (en) * | 2020-02-07 | 2021-08-12 | Nova-Tech Engineering, Llc | Shrimp processing system, processing apparatus and methods |
| US20210321593A1 (en) * | 2020-04-21 | 2021-10-21 | InnovaSea Systems, Inc. | Systems and methods for fish volume estimation, weight estimation, and analytic value generation |
| CN111832532A (en) * | 2020-07-24 | 2020-10-27 | 上海电气集团自动化工程有限公司 | Online visual detection method and system for crayfish pose identification |
| CN113537175A (en) * | 2021-09-16 | 2021-10-22 | 长沙鹏阳信息技术有限公司 | Same-fence swinery average weight estimation method based on computer vision |
| CN114399674A (en) * | 2022-01-21 | 2022-04-26 | 岭南师范学院 | Hyperspectral image technology-based shellfish toxin nondestructive rapid detection method and system |
Non-Patent Citations (2)
| Title |
|---|
| ARIF SETIAWAN等: "Underwater Shrimp Digital Image Segmentation Using Edge Detection Method on Fog Network", 《2022 2ND INTERNATIONAL CONFERENCE ON INFORMATION TECHNOLOGY AND EDUCATION (ICIT&E)》 * |
| 高升等: "红提糖度和硬度的高光谱成像无损检测", 《光学学报》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117299596A (en) * | 2023-08-14 | 2023-12-29 | 江苏秦郡机械科技有限公司 | Material screening system and method for automatic detection |
| CN117299596B (en) * | 2023-08-14 | 2024-05-24 | 江苏秦郡机械科技有限公司 | Material screening system and method for automatic detection |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114972895B (en) | 2022-09-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Mazen et al. | Ripeness classification of bananas using an artificial neural network | |
| CN111968080B (en) | Method for detecting quality of inside and outside of Feicheng peaches based on hyperspectral and deep learning | |
| Liu et al. | Design of a tomato classifier based on machine vision | |
| CN114972895B (en) | Nondestructive testing method and device for crayfish quality | |
| CN111712769A (en) | Method, apparatus, system, and program for setting lighting condition, and storage medium | |
| Grassi et al. | Fish fillet authentication by image analysis | |
| Sumaryanti et al. | Digital image based identification of rice variety using image processing and neural network | |
| Liu et al. | Rapid identification of chrysanthemum teas by computer vision and deep learning | |
| Zhang et al. | Computer vision estimation of the volume and weight of apples by using 3d reconstruction and noncontact measuring methods | |
| CN115908257A (en) | Defect recognition model training method and fruit and vegetable defect recognition method | |
| Lin et al. | A residual dense comprehensively regulated convolutional neural network to identify spectral information for egg quality traceability | |
| Setiadi et al. | Adulteration detection in minced beef using low-cost color imaging system coupled with deep neural network | |
| CN111968087A (en) | Plant disease area detection method | |
| CN117011274A (en) | Automatic glass bottle detection system and method thereof | |
| CN107024480A (en) | A kind of stereoscopic image acquisition device | |
| Thipsrirach et al. | Classification of durian maturity using a convolutional neural network. | |
| Chen et al. | Classification of rice appearance quality based on LS-SVM using machine vision | |
| CN110609011A (en) | Near-infrared hyperspectral detection method and system for starch content of single-kernel corn seeds | |
| Kwon et al. | Quality grading of Hanwoo (Korean native cattle breed) sub-images using convolutional neural network | |
| Zhao et al. | Automated phenotyping of single seeds using a novel volume sculpting framework | |
| CN114965486A (en) | Hyperspectral imaging-based slight bruise detection method and device for persimmons | |
| Septiarini et al. | Image processing for maturity classification of tomato using otsu and manhattan distance methods. | |
| Tan et al. | Automated tomato grading system using computer vision (CV) and deep neural network (DNN) algorithm | |
| Singh et al. | Rice kernels classification with deep learning using a modified dataset mimicking real-world conditions | |
| Cao et al. | A computer vision program that identifies and classifies fish species |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |