CN113640229B - Rapid nondestructive testing method and device for soybean quality with multiple parameters - Google Patents
Rapid nondestructive testing method and device for soybean quality with multiple parameters Download PDFInfo
- Publication number
- CN113640229B CN113640229B CN202110943917.5A CN202110943917A CN113640229B CN 113640229 B CN113640229 B CN 113640229B CN 202110943917 A CN202110943917 A CN 202110943917A CN 113640229 B CN113640229 B CN 113640229B
- Authority
- CN
- China
- Prior art keywords
- soybean
- dimensional
- prediction model
- hyperspectral
- quality
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- Y02A40/10—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in agriculture
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Health & Medical Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention provides a soybean quality multi-parameter rapid nondestructive testing method and a device, and the method comprises the following steps: acquiring two-dimensional gray images of a detection sample under different wave bands, and synthesizing the two-dimensional gray images under all the different wave bands into a three-dimensional hyperspectral image; segmenting the three-dimensional hyperspectral image according to each soybean, and determining hyperspectral image data of each soybean; extracting the spectral data of each soybean from the hyperspectral image data, respectively inputting the spectral data into a prediction model after preprocessing and characteristic wave band screening, and determining a plurality of quality parameter values of each soybean according to the output result of the prediction model. The method can simultaneously extract spectral images of a plurality of varieties of detection samples without damaging soybean seeds, and realize simultaneous detection of a plurality of qualities of a plurality of varieties of samples; after the two-dimensional gray scale images under all the wave bands are synthesized into the three-dimensional hyperspectral image, the information of a plurality of characteristic wave bands is extracted, and the high-accuracy detection of multiple varieties and multiple parameters can be realized.
Description
Technical Field
The invention relates to the field of nondestructive testing of agricultural products, in particular to a soybean quality multi-parameter rapid nondestructive testing method and device.
Background
Soybean is an important economic crop in the world, and the edible vegetable oil for animal nutrition and food processing industry is mainly from soybean seeds. Due to the increasing demand for energy and the limited reserves of non-renewable fossil fuels, biofuels from plants, including soy, are receiving increasing attention from the industry. In current breeding programs, the nutritional value of soybean seeds, such as sugars, proteins, oils and fatty acids, is typically measured by wet chemistry methods. The general procedure for wet chemical analysis includes seed grinding, extraction, identification and quantification using expensive instruments. Wet chemistry methods are destructive, time consuming, labor intensive and expensive, and are not suitable for quality analysis of large numbers of individual seeds in breeding programs. More importantly, the use of destructive methods will destroy seeds that are likely to have significant breeding potential during breeding, which is a challenge in early breeding stages because the number of seeds available is inherently very small.
However, the detection of crude protein content, fatty acid content and sucrose content of soybean requires complicated biochemical tests, and only one quality of soybean can be determined by one biochemical method, but a plurality of qualities of soybean cannot be determined simultaneously. Even if the soybean quality detection is realized by the hyperspectral imaging technology, the single quality of a plurality of varieties is mainly detected.
Disclosure of Invention
Aiming at the problems of high complexity and long period in the simultaneous detection of multiple qualities of multiple varieties of soybeans in the prior art, the invention provides a method and a device for rapidly detecting the quality of soybeans in a multi-parameter and nondestructive mode.
The invention provides a soybean quality multi-parameter rapid nondestructive testing method, which comprises the following steps: acquiring two-dimensional gray images of a detection sample under a plurality of wave bands, and synthesizing the two-dimensional gray images under all the wave bands into a three-dimensional hyperspectral image; segmenting the three-dimensional hyperspectral image according to each soybean, and determining hyperspectral image data of each soybean; extracting spectral data of each soybean from the hyperspectral image data, preprocessing the spectral data of each soybean, screening characteristic wave bands, inputting the spectral data of each soybean into a prediction model respectively, and determining a plurality of quality parameter values of each soybean according to an output result of the prediction model; and the prediction model is obtained by training according to the training samples with known quality parameter values by taking the quality parameter values as labels and the spectral data of each soybean of the training samples as input.
According to the soybean quality multi-parameter rapid nondestructive testing method provided by the embodiment of the invention, after the spectral data of each soybean is preprocessed and the characteristic wave band is screened, before the spectral data is respectively input into a prediction model, the method further comprises the following steps: and according to the training sample, taking a plurality of known quality parameter values as labels, and taking the spectrum data corresponding to the plurality of characteristic wave bands screened out after pretreatment as input, constructing a model, and training to obtain the prediction model.
According to the soybean quality multi-parameter rapid nondestructive testing method provided by the embodiment of the invention, before the model is constructed and trained to obtain the prediction model, the method further comprises the following steps: acquiring two-dimensional gray images of training samples in different wave bands within a preset wavelength range, and synthesizing the two-dimensional gray images of the different wave bands into a three-dimensional hyperspectral training image; segmenting the three-dimensional hyperspectral training image according to each soybean, and determining hyperspectral image training data of each soybean; acquiring spectral data of each soybean by taking the whole soybean image as an interested area; and screening different wave bands of the spectral data according to an information-free variable elimination method, a genetic algorithm or a stepwise linear regression algorithm to obtain the plurality of characteristic wave bands.
According to the soybean quality multi-parameter rapid nondestructive testing method, after the prediction model is obtained, the method further comprises the following steps: verifying the accuracy of the prediction model according to a prediction set sample with known quality parameter values; and if the accuracy does not meet the preset condition, reselecting the characteristic wave band, and training the prediction model.
According to the soybean quality multi-parameter rapid nondestructive testing method provided by the embodiment of the invention, after the three-dimensional hyperspectral image is synthesized, the method further comprises the following steps: acquiring a white spectrum reference standard calibration plate image and acquiring a full-black correction image; and performing black-and-white correction on the hyperspectral image data according to the white spectrum reference standard calibration plate image and the full-black correction image.
According to the soybean quality multi-parameter rapid nondestructive testing method, quality parameters comprise fatty acid content, crude protein content and sucrose content.
The invention also provides electronic equipment which comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor executes the program to realize the steps of the soybean quality multi-parameter rapid nondestructive testing method.
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 the steps of the method for multi-parametric, fast, non-destructive testing of soybean quality as described in any of the above.
The invention also provides a soybean quality multi-parameter rapid nondestructive testing device, which comprises a camera bellows, a mobile platform, a light source, a hyperspectral imager and the electronic equipment;
the mobile platform is used for placing a detection sample, is arranged in the dark box and can perform linear reciprocating motion in the dark box;
the light source is arranged in the dark box and used for generating spectrums with different wavelengths;
the hyperspectral imager is arranged in the dark box and used for acquiring a spectral image of a detection sample;
the electronic equipment is connected with the hyperspectral imager.
According to the soybean quality multi-parameter rapid nondestructive testing device provided by the embodiment of the invention, the hyperspectral imager is a push-broom hyperspectral imager.
According to the method and the device for rapidly detecting the soybean quality parameters in a multi-parameter mode, quality parameters are detected through hyperspectral image data and a prediction model, soybean seeds do not need to be damaged, and the detection efficiency is high. The three-dimensional hyperspectral images are synthesized through the two-dimensional gray level images and then are segmented according to each soybean, so that spectrum images can be extracted from detection samples of multiple varieties simultaneously, and the simultaneous detection of multiple internal qualities of multiple varieties of soybeans is realized. After the two-dimensional gray level images under all the wave bands are synthesized into the three-dimensional hyperspectral image, the information of a plurality of characteristic wave bands is extracted, on one hand, the detection of various parameters is facilitated, and on the other hand, the detection accuracy is improved.
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 multi-parameter rapid nondestructive testing method for soybean quality provided by the present invention;
FIG. 2 is a schematic flow chart of a multi-parameter rapid nondestructive testing method for soybean quality according to another embodiment of the present invention;
FIG. 3 is a schematic diagram of an electronic device provided by the present invention;
FIG. 4 is a schematic structural diagram of a soybean quality multi-parameter rapid nondestructive testing apparatus provided by the present invention.
In the figure: the method comprises the following steps of 1-electronic equipment, 2-data lines, 3-light sources, 4-hyperspectral imager, 5-camera bellows, 6-stepping motor controller, 7-coupler, 8-linear guide rail ball screw, 9-mobile platform and 10-detection sample.
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.
Fig. 1 is a schematic flow chart of a soybean quality multi-parameter rapid nondestructive testing method provided by the present invention, and as shown in fig. 1, the present invention provides a soybean quality multi-parameter rapid nondestructive testing method, which comprises:
and S1, acquiring two-dimensional gray images of the detection sample under a plurality of wave bands, and synthesizing the two-dimensional gray images under all the wave bands into a three-dimensional hyperspectral image.
The detection sample can be soybeans of multiple varieties, for example, 10 varieties of soybeans are detected simultaneously, the detection sample is placed on a detection platform, and numbering can be carried out in advance to distinguish various varieties. And adjusting parameters of the detection device, including the height of the push-broom hyperspectral imager with the standard lens, the focal length of the lens and the like. And acquiring two-dimensional gray images under multiple wave bands within the range of 600 nm-1000 nm by using a hyperspectral imager.
And synthesizing the two-dimensional gray level images of all the wave bands into a three-dimensional hyperspectral image. For example, two-dimensional gray scale image data under different characteristic wave bands acquired by SpecGrabber software is synthesized into three-dimensional hyperspectral image data.
And S2, segmenting the three-dimensional hyperspectral image according to each soybean, and determining hyperspectral image data of each soybean.
In S2, each soybean image in the entire hyperspectral three-dimensional image is segmented. For example, each soybean image in the whole three-dimensional image is segmented by preprocessing methods such as image segmentation and edge detection, so as to obtain hyperspectral image data of each soybean. The whole soybean can be selected as an interested area, and the spectral data is extracted from the interested area.
S3, extracting spectral data of each soybean from the hyperspectral image data, preprocessing the spectral data of each soybean, screening characteristic wave bands, inputting the spectral data of each soybean into a prediction model respectively, and determining a plurality of quality parameter values of each soybean according to the output result of the prediction model;
the hyperspectral image data are image data, and the extracted spectral data represent the relationship between each wavelength and the reflectivity and can be represented in a curve form. The preprocessing includes smoothing, spectral differentiation, etc.
And the prediction model is obtained by training according to the training samples with known quality parameter values by taking the quality parameter values as labels and spectral data of each grain of soybean of the training samples as input. I.e., the predictive model in S3, is a model that has been trained through a large number of training samples. The prediction model includes a model established by an algorithm such as partial least squares regression, and includes a model established based on a deep neural network such as a convolutional neural network.
Quality parameters include fatty acid content (including oleic acid and linoleic acid), crude protein content, and sucrose content. For example, the spectral data of each soybean of a training sample is used as input, the determined fatty acid content, crude protein content and sucrose content are used as output, and a model is constructed and trained to obtain a prediction model. And inputting the spectral data of each soybean comprising the multi-quality detection samples into the prediction model based on the obtained prediction model to obtain the fatty acid content, the crude protein content and the sucrose content of the detection samples.
According to the soybean quality multi-parameter rapid nondestructive testing method, the quality parameter is tested through the hyperspectral image data and the prediction model, the soybean seeds do not need to be damaged, and the testing efficiency is high. The three-dimensional hyperspectral images are synthesized through the two-dimensional gray level images and then are segmented according to each soybean, so that spectrum data can be extracted from detection samples of multiple varieties simultaneously, and the simultaneous detection of multiple internal qualities of multiple varieties of soybeans is realized. After the two-dimensional gray scale images under all the wave bands are synthesized into the three-dimensional hyperspectral image, the spectrum is extracted by taking the whole soybeans as the interested area, and the information of a plurality of characteristic wave bands is extracted, so that the detection of various parameters is facilitated on one hand, and the detection accuracy is improved on the other hand.
On the basis of the above embodiment, as an optional embodiment, after the spectral data of each soybean is preprocessed and characteristic band-screened, before the spectral data is respectively input into the prediction model, the method further includes: and according to the training sample, taking a plurality of known quality parameter values as labels, and taking spectrum data corresponding to a plurality of characteristic wave bands screened after pretreatment as input, constructing a model, and training to obtain the prediction model.
The test is performed by training samples before using the predictive model. The specific steps may include:
the method comprises the steps of numbering soybean seeds of 10 varieties in sequence, collecting hyperspectral data of each soybean of each variety in sequence, collecting 30 soybeans as training samples each time, and determining quality parameters of the training samples in advance through methods such as biochemical tests.
And synthesizing the two-dimensional images under all the acquired wave bands to obtain three-dimensional hyperspectral image data.
And carrying out black and white correction on the obtained three-dimensional hyperspectral data of the soybean.
And respectively obtaining the spectral data of each soybean from the corrected hyperspectral image.
And performing data preprocessing, such as smoothing, spectral differentiation and the like, on the extracted spectral data of each soybean. The spectrum data is preprocessed, so that the noise interference of the soybean spectrum data is removed, and the signal-to-noise ratio is improved.
The spectral data of each soybean is used as model input, each predetermined quality parameter is used as a label, a prediction model is constructed for training, and the prediction model is obtained for the detection of the embodiment.
The soybean quality multi-parameter rapid nondestructive testing method can simultaneously test a plurality of quality parameters of a plurality of varieties of samples to be tested through a pre-trained prediction model.
On the basis of the above embodiment, as an optional embodiment, before obtaining the prediction model, the method further includes: acquiring two-dimensional gray images of training samples in different wave bands within a preset wavelength range, and synthesizing the two-dimensional gray images in different wave bands into a three-dimensional hyperspectral training image; segmenting the three-dimensional hyperspectral training image according to each soybean, and determining hyperspectral image training data of each soybean; extracting spectral data by taking the whole soybean image as an interested area; and screening different wave bands of the spectral data according to an information-free variable elimination method, a genetic algorithm or a stepwise linear regression algorithm to obtain the plurality of characteristic wave bands.
In order to improve the detection accuracy of the model, for the plurality of characteristic bands in S1, effective wavelength points can be selected from a preset wavelength range (for example, a wavelength range of 600nm to 1000 nm) by using an information-free variable elimination method, a genetic algorithm, a stepwise linear regression algorithm, or the like, and wavelength points with low information content and no information content are removed as the characteristic bands.
And establishing a prediction model by using the screened characteristic wavelength parameters for training so as to detect the fatty acid content, the crude protein content and the sucrose content of each soybean grain of the detected sample.
On the basis of the foregoing embodiment, as an optional embodiment, after obtaining the prediction model, the method further includes: verifying the accuracy of the prediction model according to a prediction set sample with known quality parameter values; and if the accuracy does not meet the preset condition, reselecting the characteristic wave band and training the prediction model.
For the initial training sample, the training set and the prediction set are divided, and the obtained spectral data of all soybean seeds with known quality parameters are calculated according to the ratio of 3: the ratio of 1 is randomly divided into a training set and a prediction set, the training set is used for establishing a prediction model, and the prediction set is used for checking the accuracy of the established model. After the training of the above embodiment is completed, the prediction set data is imported into the established prediction model, and the accuracy of the established model is determined according to the difference between the predicted value calculated by the prediction model and the actual value measured by the actual experiment. If the preset condition is not met (if 98 percent is set), the characteristic wave band is reselected for training.
On the basis of the foregoing embodiment, as an optional embodiment, after the synthesizing the three-dimensional hyperspectral image, the method further includes: acquiring a white spectrum reference standard calibration plate image and acquiring a full-black correction image; and performing black-and-white correction on the hyperspectral image data according to the white spectrum reference standard calibration plate image and the full-black correction image. A white spectral reference standard calibration plate with a nominal reflectivity of 95% was placed in a similar position on the seed to obtain a white corrected image. The lens is then covered with a camera cover to obtain a full black corrected image. And performing black and white correction on the hyperspectral image data according to the standard calibration plate image and the all-black correction image.
According to the soybean quality multi-parameter rapid nondestructive testing method, the black and white correction is performed on the three-dimensional hyperspectral image data through the black and white correction file obtained by the adjusting instrument, so that the interference of factors such as dark current and the like can be eliminated.
Fig. 2 is a schematic flow chart of a soybean quality multi-parameter rapid non-destructive inspection method according to another embodiment of the present invention, wherein steps a-H can be taken in the above embodiments and fig. 2, which are not repeated herein. In a specific embodiment, the experimental samples are 10 varieties of soybean seeds, 300 soybean seeds are all sourced from a soybean breeding research center, prediction is carried out by the soybean quality multi-parameter nondestructive testing method, the correlation coefficient between the predicted value and the true value is larger than 0.9, and the root mean square error is small.
Fig. 3 is a schematic structural diagram of an electronic device provided in the present invention, and as shown in fig. 3, the electronic device may include: a processor (processor)301, a communication interface (communication interface)302, a memory (memory)303 and a communication bus 304, wherein the processor 301, the communication interface 302 and the memory 303 complete communication with each other through the communication bus 304. Processor 301 may invoke logic instructions in memory 303 to perform a method for fast, non-destructive testing of soybean quality multiparameters, the method comprising: acquiring two-dimensional gray images of a detection sample under a plurality of wave bands, and synthesizing the two-dimensional gray images under all the wave bands into a three-dimensional hyperspectral image; segmenting the three-dimensional hyperspectral image according to each soybean, and determining hyperspectral image data of each soybean; extracting spectral data of each soybean from the hyperspectral image data, preprocessing the spectral data of each soybean, screening characteristic wave bands, inputting the spectral data of each soybean into a prediction model respectively, and determining a plurality of quality parameter values of each soybean according to an output result of the prediction model; and the prediction model is obtained by training according to the training samples with known quality parameter values by taking the quality parameter values as labels and the spectral data of each soybean of the training samples as input.
In addition, the logic instructions in the memory 303 may be implemented in the form of software functional units and stored in a computer readable storage medium when the logic instructions 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: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
In another aspect, the present invention also provides a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions, which when executed by a computer, enable the computer to perform the method for multi-parameter fast nondestructive testing of soybean quality provided by the above methods, the method comprising: acquiring two-dimensional gray images of a detection sample under a plurality of wave bands, and synthesizing the two-dimensional gray images under all the wave bands into a three-dimensional hyperspectral image; segmenting the three-dimensional hyperspectral image according to each soybean, and determining hyperspectral image data of each soybean; extracting the spectral data of each soybean from the hyperspectral image data, respectively inputting the spectral data of each soybean into a prediction model after preprocessing and characteristic wave band screening, and determining a plurality of quality parameter values of each soybean according to the output result of the prediction model; and the prediction model is obtained by training according to the training samples with known quality parameter values by taking the quality parameter values as labels and the spectral data of each soybean of the training samples as input.
In yet 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 perform the method for multi-parameter fast nondestructive testing of soybean quality provided by the above embodiments, the method comprising: acquiring two-dimensional gray images of a detection sample under a plurality of wave bands, and synthesizing the two-dimensional gray images under all the wave bands into a three-dimensional hyperspectral image; segmenting the three-dimensional hyperspectral image according to each soybean, and determining hyperspectral image data of each soybean; extracting spectral data of each soybean from the hyperspectral image data, preprocessing the spectral data of each soybean, screening characteristic wave bands, inputting the spectral data of each soybean into a prediction model respectively, and determining a plurality of quality parameter values of each soybean according to an output result of the prediction model; and the prediction model is obtained by training according to the training samples with known quality parameter values by taking the quality parameter values as labels and spectral data of each grain of soybean of the training samples as input.
Fig. 4 is a schematic structural diagram of a soybean quality multi-parameter rapid nondestructive testing apparatus provided by the present invention, and as shown in fig. 4, the soybean quality multi-parameter rapid nondestructive testing apparatus includes: the system comprises a camera bellows 5, a mobile platform 9, a light source 3, a hyperspectral imager 4 and an electronic device 1; the moving platform 9 is used for placing a detection sample 10, is arranged in the dark box 5 and can perform linear reciprocating movement in the dark box 5; the light source 3 is arranged in the dark box 5 and is used for generating spectrums with different wavelengths; the hyperspectral imager 4 is arranged in the dark box 5 and is used for acquiring a spectral image of the detection sample 10; the electronic device 1 is connected (e.g. via a data line 2) to the hyperspectral imager 4.
On the basis of the above embodiment, as an optional embodiment, the hyperspectral imager is a push-broom hyperspectral imager.
Particularly, install light source 3 and the push-broom formula hyperspectral imager 4 that has the standard camera lens at camera bellows 5 top, camera bellows 5 bottom sets up moving platform 9, and moving platform 9 mobilizable setting is under light source 3 and hyperspectral imager 4 for place test sample 10, and moving platform 9 accessible helicitic texture is connected with linear guide rail ball 8, and step motor controller 6 passes through shaft coupling 7 and is connected with linear guide rail ball 8, and shaft coupling 7, linear guide rail ball 8, moving platform 9 all set up the interior bottom of camera bellows 5, step motor controller 6 sets up in camera bellows 5 outside, and step motor controller 6 controls linear guide rail ball 8 through shaft coupling 7 and removes to drive moving platform 9 parallel movement. The soybean sample 10 is laid on the moving platform 9.
The electronic device 1 may install specribbon software for configuring camera settings, setting imaging rates and controlling data acquisition. The data line 2 adopts a USB 3.0 interface and has the function of connecting the electronic equipment 1 and the push-broom hyperspectral imager 4 with a standard lens. The push-broom hyperspectral imager 4 with the standard lens can adopt an onboard hyperspectral camera, the model is OCI-UAV-1000, the data acquisition mode is a push-broom type, and the spectral range is 600 nm-1000 nm; the size of a camera lens (35 mm fixed focus, 18-degree field angle) is 80 mm x 60 mm x 60 mm; the function of the method is to shoot high-quality soybean hyperspectral images.
The camera bellows 5 is a rectangular closed box built by section bars with certain strength and rigidity, and the box can be sealed by a completely black wood board. The function is to install other functional components, and the interference of the surrounding environment when the reflection hyperspectral data of the detection sample 10 is collected is avoided.
The stepper motor controller 6 can be a NEMA 23 model, the length of the machine body is 56 mm, the current is 2.8A and is 1.1N.m, the shaft diameter is 6.35 mm, and the motion mode is single-step and continuous. The function is to control the moving speed, the advancing, the retreating, the moving distance and the like of the screw rod type moving platform by controlling the forward and reverse rotation and the rotation time of the screw rod type moving platform.
The coupler 7 is a common component, and functions to connect the stepping motor controller 6 and the linear guide ball screw 8, so that the linear guide ball screw 8 rotates.
The linear guide ball screw 8 is a common component, and functions to control the linear advance and retreat of the mobile platform.
The movable platform 9 can be a rectangular steel plate, and has the function of placing the detection samples, so that the detection samples can be flatly laid on the steel plate, and the steel plate can stably move forwards and backwards in the moving process. Meanwhile, when the image is subjected to black and white correction, a white spectrum reference standard calibration plate can be placed.
According to the soybean quality multi-parameter rapid nondestructive testing device provided by the embodiment of the invention, the linear guide rail ball screw is controlled to move through the coupler so as to drive the moving platform to move in parallel, so that the moving platform can be adjusted within the range of a lens of a hyperspectral imager, and a plurality of varieties of test samples can be placed in combination with the moving platform, so that the simultaneous testing of a plurality of varieties and a plurality of quality parameters can be realized.
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 can be implemented by software plus a necessary general hardware platform, and certainly can 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, and 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 (7)
1. A soybean quality multi-parameter rapid nondestructive testing method is characterized by comprising the following steps:
acquiring two-dimensional gray images of a detection sample under multiple wave bands within the range of 600 nm-1000 nm, and synthesizing the two-dimensional gray images under all the wave bands into a three-dimensional hyperspectral image;
segmenting the three-dimensional hyperspectral image according to each soybean, and determining hyperspectral image data of each soybean;
extracting spectral data of each soybean from the hyperspectral image data, preprocessing the spectral data of each soybean, screening characteristic wave bands, inputting the spectral data of each soybean into a prediction model respectively, and determining a plurality of quality parameter values of each soybean according to an output result of the prediction model;
the prediction model is obtained by training according to the training samples with known quality parameter values by taking the quality parameter values as labels and the spectral data of each soybean of the training samples as input; the quality parameters comprise fatty acid content, crude protein content and sucrose content;
after the spectral data of each soybean is preprocessed and characteristic wave band is screened, before the spectral data is respectively input into a prediction model, the method further comprises the following steps:
according to a training sample, taking a plurality of known quality parameter values as labels, taking spectrum data corresponding to a plurality of characteristic wave bands screened after pretreatment as input, and constructing a model for training to obtain the prediction model;
before the model is trained to obtain the prediction model, the method further comprises the following steps:
acquiring two-dimensional gray images of training samples in different wave bands within a wavelength range of 600 nm-1000 nm, and synthesizing the two-dimensional gray images in different wave bands into a three-dimensional hyperspectral training image;
segmenting the three-dimensional hyperspectral training image according to each soybean, and determining hyperspectral image training data of each soybean;
extracting spectral data by taking the whole soybean image as an interested area;
and screening different wave bands of the spectral data according to an information-free variable elimination method, a genetic algorithm or a stepwise linear regression algorithm to obtain the plurality of characteristic wave bands.
2. The method of claim 1, wherein after obtaining the prediction model, the method further comprises:
verifying the accuracy of the prediction model according to a prediction set sample with known quality parameter values;
and if the accuracy does not meet the preset condition, reselecting the characteristic wavelength and training the prediction model.
3. The soybean quality multi-parameter rapid nondestructive testing method according to any one of claims 1 to 2, wherein after the three-dimensional hyperspectral image is synthesized, the method further comprises the following steps:
acquiring a white spectrum reference standard calibration plate image and acquiring a full-black correction image;
and performing black-and-white correction on the hyperspectral image data according to the white spectrum reference standard calibration plate image and the full-black correction image.
4. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps of the method for multi-parametric, fast, non-destructive testing of soybean quality as claimed in any one of claims 1 to 3.
5. A non-transitory computer readable storage medium, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the steps of the method for multi-parameter fast non-destructive testing of soybean quality according to any one of claims 1 to 3.
6. A soybean quality multi-parameter rapid nondestructive testing device is characterized by comprising a camera bellows, a mobile platform, a light source, a hyperspectral imager and the electronic equipment of claim 4;
the mobile platform is used for placing a detection sample, is arranged in the dark box and can perform linear reciprocating motion in the dark box;
the light source is arranged in the dark box and used for generating spectrums with different wavelengths;
the hyperspectral imager is arranged in the dark box and used for acquiring a spectral image of a detection sample;
the electronic equipment is connected with the hyperspectral imager.
7. The multi-parameter rapid nondestructive testing apparatus for soybean quality according to claim 6, wherein said hyperspectral imager is a push-and-scan hyperspectral imager.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110943917.5A CN113640229B (en) | 2021-08-17 | 2021-08-17 | Rapid nondestructive testing method and device for soybean quality with multiple parameters |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110943917.5A CN113640229B (en) | 2021-08-17 | 2021-08-17 | Rapid nondestructive testing method and device for soybean quality with multiple parameters |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113640229A CN113640229A (en) | 2021-11-12 |
| CN113640229B true CN113640229B (en) | 2022-09-06 |
Family
ID=78422547
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110943917.5A Active CN113640229B (en) | 2021-08-17 | 2021-08-17 | Rapid nondestructive testing method and device for soybean quality with multiple parameters |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113640229B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117007552B (en) * | 2023-10-07 | 2024-02-06 | 北京市农林科学院智能装备技术研究中心 | Watermelon maturity detection method, device, system, electronic equipment and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104406940B (en) * | 2014-12-01 | 2017-05-03 | 上海理工大学 | Nondestructive identification method of rice with industrial wax |
-
2021
- 2021-08-17 CN CN202110943917.5A patent/CN113640229B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113640229A (en) | 2021-11-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | Detection of early rottenness on apples by using hyperspectral imaging combined with spectral analysis and image processing | |
| CN106525732B (en) | Rapid nondestructive detection method for internal and external quality of apple based on hyperspectral imaging technology | |
| CN103822879A (en) | Nondestructive detection method of swelled kiwi fruits based on hyperspectral imaging technology | |
| CN104406940B (en) | Nondestructive identification method of rice with industrial wax | |
| CN110849828A (en) | Saffron crocus classification method based on hyperspectral image technology | |
| CN106355579A (en) | Defect detecting method of cigarette carton surface wrinkles | |
| CN111380813B (en) | Portable wheat seed multi-quality nondestructive testing device and testing method | |
| CN111044483A (en) | Method, system and medium for determining pigment in cream based on near infrared spectrum | |
| CN110596117A (en) | Hyperspectral imaging-based rapid nondestructive detection method for apple surface damage | |
| Rattanaphongphak et al. | Design of machine vision system for sugarcane buds or rings detection | |
| CN105021617A (en) | Hyperspectral imaging-based measuring equipment of chlorophyll content of whole rice plant and method thereof | |
| CN113640229B (en) | Rapid nondestructive testing method and device for soybean quality with multiple parameters | |
| Uluişik et al. | Image processing based machine vision system for tomato volume estimation | |
| Zhihui et al. | Nondestructive detection of infertile hatching eggs based on spectral and imaging information | |
| CN114419311B (en) | Multi-source information-based passion fruit maturity nondestructive testing method and device | |
| CN103528967A (en) | Hyperspectral image based overripe Lonicera edulis fruit identification method | |
| CN107576600B (en) | Quick detection method for matcha granularity grade | |
| CN104089925A (en) | Hyperspectral imaging-based Target area extraction method for detecting shrimp quality | |
| CN116297236A (en) | Method and device for identifying vitality of single corn seeds based on hyperspectrum | |
| CN110929787A (en) | Apple objective grading system based on images | |
| CN114136920A (en) | Hyperspectrum-based single-grain hybrid rice seed variety identification method | |
| CN106018292A (en) | Non-destructive testing device for protein conformation in egg white and method of non-destructive testing device | |
| CN110837823A (en) | Method for generating seed variety identification model, identification method and device | |
| Shah et al. | Automatically localising ROIs in hyperspectral images using background subtraction techniques | |
| CN115372272A (en) | Apple damage detection method |
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 |