A rice quality assessment method and system

Abstract

This invention provides a method and system for assessing rice quality. The invention relates to the field of rice quality assessment technology. The method includes: determining the average vibration amplitude and average vibration frequency of each grain of rice using a vibration video processing model based on the up-and-down vibration video of a pre-set weight of rice to be tested; determining the vibration consistency based on the average vibration amplitude and average vibration frequency of each grain of rice; determining whether the vibration consistency is greater than or equal to a vibration consistency threshold; if the vibration consistency is less than the vibration consistency threshold, determining that the quality of the rice to be tested is unqualified; if the vibration consistency is greater than or equal to the vibration consistency threshold, controlling an ultrasonic transmitter to emit ultrasonic signals to the rice to be tested and receiving the ultrasonic signals returned by the rice; and using a sound wave analysis model based on the ultrasonic signals returned by the rice to determine whether the quality of the rice to be tested is qualified. This method can quickly and accurately assess the quality of rice.

Description

Technical Field

[0001] The invention relates to the technical field of rice quality assessment, and in particular to a rice quality assessment method and system. Background Art

[0002] During rice production and processing, quality assessment is a key link to ensure product quality and consumer satisfaction. Traditional rice quality assessment methods mainly rely on manual visual inspection, which is not only inefficient but also easily affected by the subjective judgment of inspectors, resulting in inconsistency and inaccuracy in the assessment results. Existing automated rice quality detection technologies often require the use of complex testing equipment and long testing cycles when processing a large number of samples, which is not efficient enough in large-scale testing. Especially in the grain trade and export links. Therefore, rapid and accurate rice quality assessment is of great significance for reducing costs and improving competitiveness.

[0003] Therefore, how to quickly and accurately evaluate the quality of rice is an urgent problem to be solved. Summary of the invention

[0004] The main technical problem solved by the invention is how to quickly and accurately evaluate the quality of rice.

[0005] According to a first aspect, the present invention provides a rice quality assessment method, comprising: obtaining an up-and-down vibration video of rice to be detected of a preset weight, wherein the up-and-down vibration video of the rice to be detected of the preset weight is obtained by being shot by a high-speed camera; determining an average vibration amplitude and an average vibration frequency of each grain of rice based on the up-and-down vibration video of the rice to be detected of the preset weight using a vibration video processing model; determining a vibration consistency based on the average vibration amplitude and the average vibration frequency of each grain of rice; judging whether the vibration consistency is greater than or equal to a vibration consistency threshold; if the vibration consistency is less than the vibration consistency threshold, determining that the quality of the rice to be detected is unqualified, and if the vibration consistency is greater than or equal to the vibration consistency threshold, controlling an ultrasonic transmitter to transmit an ultrasonic signal to the rice to be detected and receiving an ultrasonic signal returned by the rice to be detected; and determining whether the quality of the rice to be detected is qualified based on the ultrasonic signal returned by the rice to be detected using an acoustic wave analysis model.

[0006] Furthermore, the acoustic wave analysis model is an artificial neural network model, the input of the acoustic wave analysis model is the ultrasonic signal returned by the rice to be tested, and the output of the acoustic wave analysis model is whether the quality of the rice to be tested is qualified or unqualified.

[0007] Furthermore, the method also includes: obtaining a heating video of the rice to be detected and electronic nose data during the heating process of the rice to be detected; determining the cracking degree of the rice to be detected and the aroma degree of the rice to be detected using a heating data processing model based on the heating video of the rice to be detected and the electronic nose data during the heating process of the rice to be detected; if the cracking degree of the rice to be detected is greater than a cracking threshold, determining that the quality of the rice to be detected is unqualified; if the aroma degree of the rice to be detected is less than an aroma threshold, determining that the quality of the rice to be detected is unqualified.

[0008] Furthermore, the vibration video processing model is a gated loop unit, the input of the vibration video processing model is the up and down vibration video of the rice to be tested with the preset weight, and the output of the vibration video processing model is the average vibration amplitude and average vibration frequency of each grain of rice.

[0009] Furthermore, the heating data processing model is a gated loop unit, the input of the heating data processing model is the heating video of the rice to be detected and the electronic nose data during the heating process of the rice to be detected, and the output of the heating data processing model is the cracking degree of the rice to be detected and the aroma degree of the rice to be detected.

[0010] According to a second aspect, the present invention provides a rice quality assessment system, comprising: an acquisition module, used to acquire a video of up and down vibration of rice to be detected with a preset weight, wherein the video of up and down vibration of rice to be detected with a preset weight is obtained by shooting with a high-speed camera;

[0011] A vibration video processing module, used for determining an average vibration amplitude and an average vibration frequency of each grain of rice based on the up and down vibration video of the rice to be detected with the preset weight using a vibration video processing model;

[0012] a vibration consistency determination module, configured to determine the vibration consistency based on the average vibration amplitude and the average vibration frequency of each grain of rice;

[0013] A judging module, used to judge whether the vibration consistency is greater than or equal to a vibration consistency threshold;

[0014] A transmitting module, configured to determine that the quality of the rice to be tested is unqualified if the vibration consistency is less than a vibration consistency threshold, and to control an ultrasonic transmitter to transmit an ultrasonic signal to the rice to be tested and receive an ultrasonic signal returned by the rice to be tested if the vibration consistency is greater than or equal to the vibration consistency threshold;

[0015] The quality determination module is used to determine whether the quality of the rice to be detected is qualified by using a sound wave analysis model based on the ultrasonic signal returned by the rice to be detected.

[0016] Furthermore, the acoustic wave analysis model is an artificial neural network model, the input of the acoustic wave analysis model is the ultrasonic signal returned by the rice to be tested, and the output of the acoustic wave analysis model is whether the quality of the rice to be tested is qualified or unqualified.

[0017] Furthermore, the system is also used for:

[0018] Obtaining a heating video of the rice to be tested and electronic nose data during the heating process of the rice to be tested;

[0019] Determine the cracking degree of the rice to be detected and the aroma degree of the rice to be detected using a heating data processing model based on the heating video of the rice to be detected and the electronic nose data during the heating process of the rice to be detected;

[0020] If the cracking degree of the rice to be tested is greater than the cracking threshold, it is determined that the quality of the rice to be tested is unqualified;

[0021] If the aroma degree of the rice to be detected is less than the aroma threshold, it is determined that the quality of the rice to be detected is unqualified.

[0022] Furthermore, the vibration video processing model is a gated loop unit, the input of the vibration video processing model is the up and down vibration video of the rice to be tested with the preset weight, and the output of the vibration video processing model is the average vibration amplitude and average vibration frequency of each grain of rice.

[0023] Furthermore, the heating data processing model is a gated loop unit, the input of the heating data processing model is the heating video of the rice to be detected and the electronic nose data during the heating process of the rice to be detected, and the output of the heating data processing model is the cracking degree of the rice to be detected and the aroma degree of the rice to be detected.

[0024] The present invention provides a rice quality assessment method and system. The method comprises the following steps: obtaining a video of up and down vibration of rice to be detected with a preset weight, wherein the video of up and down vibration of the rice to be detected with a preset weight is obtained by shooting with a high-speed camera; determining an average vibration amplitude and an average vibration frequency of each grain of rice based on the video of up and down vibration of the rice to be detected with a vibration video processing model; determining vibration consistency based on the average vibration amplitude and the average vibration frequency of each grain of rice; judging whether the vibration consistency is greater than or equal to a vibration consistency threshold; if the vibration consistency is less than the vibration consistency threshold, determining that the quality of the rice to be detected is unqualified; if the vibration consistency is greater than or equal to the vibration consistency threshold, controlling an ultrasonic transmitter to transmit an ultrasonic signal to the rice to be detected and receiving an ultrasonic signal returned by the rice to be detected; and determining whether the quality of the rice to be detected is qualified based on the ultrasonic signal returned by the rice to be detected using an acoustic wave analysis model. The method can quickly and accurately assess the quality of rice. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A schematic diagram of a flow chart of a rice quality assessment method provided by an embodiment of the present invention;

[0026] Figure 2 A schematic diagram of a process for using a heating data processing model to determine whether the quality of rice to be tested is qualified according to an embodiment of the present invention;

[0027] Figure 3 A schematic diagram of a rice quality assessment system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In an embodiment of the present invention, there is provided Figure 1 A rice quality assessment method is shown, the rice quality assessment method comprising steps S1 to S6:

[0029] Step S1, obtaining a video of up and down vibration of the rice to be tested with a preset weight, wherein the video of up and down vibration of the rice to be tested with a preset weight is obtained by shooting with a high-speed camera.

[0030] The preset weight refers to a predetermined weight of the rice to be tested, for example, the preset weight of the rice to be tested is 1 gram, 2 grams, 3 grams, 5 grams, etc.

[0031] Rice to be tested refers to rice samples that need to be evaluated for quality.

[0032] The up and down vibration video of the rice to be tested is a video captured by a high-speed camera after the rice to be tested is vibrated up and down regularly using a vibration device.

[0033] Step S2, based on the up and down vibration video of the rice to be tested with the preset weight, using a vibration video processing model to determine the average vibration amplitude and average vibration frequency of each grain of rice.

[0034] The vibration video processing model is a gated loop unit, the input of the vibration video processing model is the up and down vibration video of the rice to be detected with the preset weight, and the output of the vibration video processing model is the average vibration amplitude and average vibration frequency of each grain of rice.

[0035] The gated recurrent unit model includes a gated recurrent unit (GRU). The gated recurrent unit model controls the flow of information by introducing a gating mechanism. The gated recurrent unit model introduces two key gating units, an update gate and a reset gate, to dynamically adjust the importance of information through learning. The update gate is used to control the weight between the hidden state of the previous moment and the current input, and the reset gate is used to control whether to reset the historical information to the default value. These two gating units enable the gated recurrent unit model to adaptively choose to retain old information and obtain new information, and better capture the short-term and long-term dependencies in the input sequence. By using the gated recurrent unit model, the sequence data of the up and down vibration video of the preset weight of the rice to be tested can be effectively processed, and the dependencies between different time steps can be captured, so as to determine the average vibration amplitude and average vibration frequency of each grain of rice. In some embodiments, the vibration video processing model can be obtained by training the training samples by the gradient descent method.

[0036] In some embodiments, the vibration video processing model includes a separation layer, an amplitude determination layer, and a frequency determination layer. The separation layer, the amplitude determination layer, and the frequency determination layer all include a gated loop unit. The input of the separation layer is the up and down vibration video of the rice to be detected of the preset weight, the output of the separation layer is the up and down vibration segmented video of each grain of rice, the input of the amplitude determination layer is the up and down vibration segmented video of each grain of rice, the output of the amplitude determination layer is the average vibration amplitude of each grain of rice, the input of the frequency determination layer is the up and down vibration segmented video of each grain of rice, and the output of the frequency determination layer is the average vibration frequency of each grain of rice. Through the separation layer, the vibration video of each grain of rice can be extracted separately, which is convenient for the subsequent determination of the amplitude and frequency. The amplitude determination layer determines the average vibration amplitude of each grain of rice. The frequency determination layer determines the average vibration frequency of each grain of rice.

[0037] In some embodiments, the amplitude determination layer includes a peak image determination sublayer, a valley image determination sublayer, a peak displacement determination sublayer, a valley displacement determination sublayer, and an average amplitude calculation sublayer. The input of the peak image determination sublayer is a segmented video of the up and down vibration of each grain of rice, the output of the peak image determination sublayer is a peak image of each up and down vibration of each grain of rice, the input of the valley image determination sublayer is a segmented video of the up and down vibration of each grain of rice, the output of the valley image determination sublayer is a valley image of each up and down vibration of each grain of rice, the input of the peak displacement determination sublayer is a peak image of each up and down vibration of each grain of rice, the output of the peak displacement determination sublayer is a peak value of each up and down vibration of each grain of rice, the input of the valley displacement determination sublayer is a valley image of each up and down vibration of each grain of rice, the output of the valley displacement determination sublayer is a valley value of each up and down vibration of each grain of rice, the input of the average amplitude calculation sublayer is the peak value of each up and down vibration of each grain of rice and the valley value of each up and down vibration of each grain of rice, and the output of the average amplitude calculation sublayer is the average vibration amplitude of each grain of rice. The peak image determination sublayer and the valley image determination sublayer are responsible for extracting the peak and valley images of each up and down vibration of each grain of rice. The peak displacement determination sublayer and the valley displacement determination sublayer are used to determine the peak and valley displacement of each up and down vibration of each grain of rice. The input of the average amplitude calculation sublayer is the peak and valley of each up and down vibration of each grain of rice, and the output is the average vibration amplitude of each grain of rice. By calculating the average of the peak and valley values, the average vibration amplitude of each grain of rice can be obtained.

[0038] By decomposing the vibration video processing model into multiple layers, the vibration information of rice can be gradually extracted and calculated, so as to better understand and analyze the vibration data. This hierarchical design can improve the interpretability, controllability and flexibility of the model.

[0039] Step S3, determining the vibration consistency based on the average vibration amplitude and average vibration frequency of each grain of rice.

[0040] Vibration consistency refers to the degree of consistency of each grain of rice in its up and down vibration. The higher the vibration consistency, the more similar each grain of rice is in vibration, which means that the weight, specifications and shape of each grain of rice are more consistent, indicating that the quality of the rice is more stable.

[0041] If the average vibration amplitude and average vibration frequency of the tested rice tend to be consistent, it means that the physical and mechanical properties of the tested rice are relatively similar and the quality is relatively stable; conversely, if the difference is large, it means that the physical and mechanical properties of each grain of rice are quite different, the quality of the rice is unstable, and there may be rice of unqualified quality.

[0042] In some embodiments, the vibration consistency can be determined by a consistency determination model. The consistency determination model is an artificial neural network model. The input of the consistency determination model is the average vibration amplitude and average vibration frequency of each grain of rice, and the output of the consistency determination model is the vibration consistency. The artificial neural network (ANN) model is a mathematical model that simulates the structure and function of the human brain nervous system. The basic structure of the artificial neural network model includes an input layer, a hidden layer, and an output layer. The input layer receives external input data, the hidden layer is responsible for processing and processing the input data, and finally passes the result to the output layer, and the output layer generates the final prediction or classification result. Artificial neural networks include feedforward neural networks, recurrent neural networks, convolutional neural networks, etc. The artificial neural network model can learn the relationship between input features and output targets through training. When determining the vibration consistency of rice, the average vibration amplitude and average vibration frequency of rice can be used as input features, and the vibration consistency can be trained as the output target. Through a large amount of sample data, the artificial neural network can learn the nonlinear relationship between the vibration amplitude and frequency and the vibration consistency. The artificial neural network gradually adjusts the connection weights to minimize the prediction error between the input features and the output targets. In this way, after the training is completed, the artificial neural network can predict the vibration consistency based on the average vibration amplitude and average vibration frequency of each grain of rice.

[0043] Step S4, determining whether the vibration consistency is greater than or equal to a vibration consistency threshold.

[0044] The vibration consistency threshold is a threshold set manually in advance.

[0045] Step S5, if the vibration consistency is less than the vibration consistency threshold, it is determined that the quality of the rice to be tested is unqualified; if the vibration consistency is greater than or equal to the vibration consistency threshold, the ultrasonic transmitter is controlled to transmit an ultrasonic signal to the rice to be tested and receive the ultrasonic signal returned by the rice to be tested.

[0046] The ultrasonic transmitter is a device for transmitting ultrasonic signals. The ultrasonic transmitter is used to send ultrasonic signals to the rice to be tested. The sound wave signal transmitted by the ultrasonic transmitter to the rice to be tested and returned by it. By analyzing the returned sound wave signal, the quality information of the rice to be tested can be obtained.

[0047] Step S6, determining whether the quality of the rice to be tested is qualified by using a sound wave analysis model based on the ultrasonic signal returned by the rice to be tested.

[0048] The acoustic wave analysis model is an artificial neural network model, the input of the acoustic wave analysis model is the ultrasonic signal returned by the rice to be detected, and the output of the acoustic wave analysis model is whether the quality of the rice to be detected is qualified or unqualified.

[0049] Ultrasonic signals can be reflected or absorbed by the density differences between different materials, thereby revealing the density distribution inside the rice. Density is one of the important indicators for evaluating rice quality, and different density distributions represent different quality characteristics. By analyzing ultrasonic signals, the density distribution of the internal structure of rice can be inferred, and then the quality of rice can be evaluated.

[0050] When ultrasonic waves pass through materials, they are affected by the internal structure of the material, such as cracks, pores, and impurities. By analyzing the attenuation of ultrasonic signals and the reflected images, the integrity and uniformity of the internal structure of rice can be inferred, and the quality of rice can be evaluated.

[0051] Since ultrasound propagates at different speeds in different media, the propagation time and intensity of the ultrasonic signal can reflect the humidity inside the rice. Humidity is an important factor affecting the quality of rice storage and processing, so the humidity changes of rice can be monitored through ultrasonic signals.

[0052] Ultrasonic signals can help detect whether there are foreign objects in the rice, such as stones, metal pieces, etc. These foreign objects may have a serious impact on the quality and safety of the rice, so ultrasonic signals can be used to detect and remove these foreign objects in a timely manner.

[0053] The ultrasonic signal returned by the rice to be tested contains rich quality information, including density, structural integrity, humidity and foreign matter detection. The acoustic wave analysis model can comprehensively and multi-angle evaluate the quality of rice by analyzing the ultrasonic signal returned by the rice to be tested. The artificial neural network model can learn the correlation between different features through training and extract features that are useful for rice quality judgment. In ultrasonic signal analysis, the model can learn the features related to rice quality in the ultrasonic signal, such as density, structural integrity, humidity, etc., so as to accurately judge whether the quality of rice is qualified.

[0054] In some embodiments, a heating data processing model can also be used to determine whether the quality of the rice to be tested is qualified. Figure 2 A schematic flow chart of using a heating data processing model to determine whether the quality of rice to be detected is qualified is provided in an embodiment of the present invention, wherein the method of using a heating data processing model to determine whether the quality of rice to be detected is qualified includes steps S21 to S24:

[0055] Step S21, obtaining a heating video of the rice to be detected and electronic nose data during the heating process of the rice to be detected.

[0056] The heating video of the rice to be tested refers to video data recording the heating process of the rice to be tested. The electronic nose data during the heating process of the rice to be tested refers to aroma data generated during the heating process of the rice to be tested obtained using an electronic nose device.

[0057] In some embodiments, a high temperature resistant camera and an electronic nose may be installed in the heating device. The electronic nose is an electronic device that simulates the human olfactory system and is used to detect and identify odors. The electronic nose consists of a gas sensor array, a signal processing unit, and a pattern recognition algorithm.

[0058] Step S22, determining the cracking degree of the rice to be detected and the aroma degree of the rice to be detected by using a heating data processing model based on the heating video of the rice to be detected and the electronic nose data during the heating process of the rice to be detected.

[0059] The heating data processing model is a gated loop unit, the input of the heating data processing model is the heating video of the rice to be detected and the electronic nose data during the heating process of the rice to be detected, and the output of the heating data processing model is the cracking degree of the rice to be detected and the aroma degree of the rice to be detected.

[0060] The gating mechanism inside the gated recurrent unit can help the model handle long-term dependencies, that is, the influence and dependency between the previous and next states over a longer time span. This is very important for judging the degree of rice cracking, because cracking is usually a process over a relatively long time scale, and the correlation between the previous and next states needs to be considered. The heating video of the rice to be tested can be regarded as a time series, which records the state changes of the rice during the heating process, such as temperature changes, cracking, etc. The gated recurrent unit can model these time series data, capture the rules and characteristics therein, and output the degree of cracking of the rice to be tested.

[0061] The electronic nose data records the odor information emitted by the rice during the heating process, and these odor characteristics are time-series. The gated recurrent unit can model the electronic nose data and capture the odor changes at different time points. The gated recurrent unit can model the electronic nose data and capture the odor changes at different time points, thereby outputting the aroma level of the rice to be tested.

[0062] Step S23: If the cracking degree of the rice to be tested is greater than the cracking threshold, it is determined that the quality of the rice to be tested is unqualified.

[0063] In this step, the degree of cracking of the rice to be tested obtained in the previous step is compared with a pre-set cracking threshold value to determine whether the quality of the rice to be tested is qualified. If the degree of cracking of the rice to be tested is greater than the cracking threshold value, it means that the rice is severely cracked during the heating process, and the quality of the rice to be tested is unqualified. By setting the cracking threshold value and making a judgment, it is possible to quickly and accurately determine whether the quality of the rice to be tested is qualified, thereby improving the efficiency of quality control. The cracking threshold value can be set in advance.

[0064] Step S24: if the aroma level of the rice to be detected is less than the aroma threshold, it is determined that the quality of the rice to be detected is unqualified.

[0065] In this step, the aroma level of the rice to be tested obtained in the previous step is compared with a preset aroma threshold to determine whether the quality of the rice to be tested is qualified. The aroma threshold can be set in advance.

[0066] Based on the same inventive concept, Figure 3 A schematic diagram of a rice quality assessment system provided by an embodiment of the present invention, the rice quality assessment system comprising:

[0067] An acquisition module 31 is used to acquire a video of the up and down vibration of the rice to be tested with a preset weight, wherein the video of the up and down vibration of the rice to be tested with a preset weight is obtained by shooting with a high-speed camera;

[0068] A vibration video processing module 32, configured to determine an average vibration amplitude and an average vibration frequency of each grain of rice based on the up and down vibration video of the rice to be detected with the preset weight using a vibration video processing model;

[0069] a vibration consistency determination module 33, configured to determine the vibration consistency based on the average vibration amplitude and the average vibration frequency of each grain of rice;

[0070] A judging module 34 is used to judge whether the vibration consistency is greater than or equal to a vibration consistency threshold;

[0071] The transmitting module 35 is used to determine that the quality of the rice to be tested is unqualified if the vibration consistency is less than the vibration consistency threshold, and to control the ultrasonic transmitter to transmit an ultrasonic signal to the rice to be tested and receive the ultrasonic signal returned by the rice to be tested if the vibration consistency is greater than or equal to the vibration consistency threshold;

[0072] The quality determination module 36 is used to determine whether the quality of the rice to be tested is qualified by using a sound wave analysis model based on the ultrasonic signal returned by the rice to be tested.

Claims

1. A rice quality assessment method, characterized in that: include: Obtaining a video of the up-and-down vibration of the rice to be tested of a preset weight, wherein the video of the up-and-down vibration of the rice to be tested of the preset weight is obtained by shooting with a high-speed camera; Based on the up and down vibration video of the rice to be detected of the preset weight, the average vibration amplitude and the average vibration frequency of each grain of rice are determined using a vibration video processing model, the vibration video processing model includes a separation layer, an amplitude determination layer, and a frequency determination layer, the separation layer, the amplitude determination layer, and the frequency determination layer all include a gated cycle unit, the input of the separation layer is the up and down vibration video of the rice to be detected of the preset weight, the output of the separation layer is the up and down vibration segmented video of each grain of rice, the input of the amplitude determination layer is the up and down vibration segmented video of each grain of rice, the output of the amplitude determination layer is the average vibration amplitude of each grain of rice, the input of the frequency determination layer is the up and down vibration segmented video of each grain of rice, and the output of the frequency determination layer is the average vibration frequency of each grain of rice; Determining the vibration consistency based on the average vibration amplitude and the average vibration frequency of each grain of rice, the vibration consistency indicates the consistency of each grain of rice in the up and down vibrations, the higher the vibration consistency, the higher the similarity of each grain of rice in vibration, the more consistent the weight, specification and shape of each grain of rice, and the more stable the quality of the rice; Determining whether the vibration consistency is greater than or equal to a vibration consistency threshold; If the vibration consistency is less than the vibration consistency threshold, it is determined that the quality of the rice to be tested is unqualified; if the vibration consistency is greater than or equal to the vibration consistency threshold, the ultrasonic transmitter is controlled to transmit an ultrasonic signal to the rice to be tested and receive the ultrasonic signal returned by the rice to be tested; Determine whether the quality of the rice to be detected is qualified by using a sound wave analysis model based on the ultrasonic signal returned by the rice to be detected; The method further comprises: Obtaining a heating video of the rice to be tested and electronic nose data during the heating process of the rice to be tested; Determine the cracking degree of the rice to be detected and the aroma degree of the rice to be detected using a heating data processing model based on the heating video of the rice to be detected and the electronic nose data during the heating process of the rice to be detected; If the cracking degree of the rice to be tested is greater than the cracking threshold, it is determined that the quality of the rice to be tested is unqualified; If the aroma degree of the rice to be detected is less than the aroma threshold, it is determined that the quality of the rice to be detected is unqualified.

2. The rice quality assessment method according to claim 1, wherein: The acoustic wave analysis model is an artificial neural network model, the input of the acoustic wave analysis model is the ultrasonic signal returned by the rice to be detected, and the output of the acoustic wave analysis model is whether the quality of the rice to be detected is qualified or unqualified.

3. The rice quality assessment method according to claim 1, wherein: The heating data processing model is a gated loop unit, the input of the heating data processing model is the heating video of the rice to be detected and the electronic nose data during the heating process of the rice to be detected, and the output of the heating data processing model is the cracking degree of the rice to be detected and the aroma degree of the rice to be detected.

4. A rice quality assessment system, characterized in that: include: An acquisition module is used to acquire a video of up and down vibration of the rice to be detected with a preset weight, wherein the video of up and down vibration of the rice to be detected with a preset weight is obtained by shooting with a high-speed camera; a vibration video processing module, for determining an average vibration amplitude and an average vibration frequency of each grain of rice based on the up and down vibration video of the rice to be detected of the preset weight using a vibration video processing model, wherein the vibration video processing model comprises a separation layer, an amplitude determination layer, and a frequency determination layer, wherein the separation layer, the amplitude determination layer, and the frequency determination layer all comprise a gated circulation unit, wherein the input of the separation layer is the up and down vibration video of the rice to be detected of the preset weight, and the output of the separation layer is the up and down vibration segmented video of each grain of rice, the input of the amplitude determination layer is the up and down vibration segmented video of each grain of rice, the output of the amplitude determination layer is the average vibration amplitude of each grain of rice, the input of the frequency determination layer is the up and down vibration segmented video of each grain of rice, and the output of the frequency determination layer is the average vibration frequency of each grain of rice; a vibration consistency determination module, for determining the vibration consistency based on the average vibration amplitude and the average vibration frequency of each grain of rice, wherein the vibration consistency indicates the degree of consistency of each grain of rice in terms of up and down vibration, and the higher the vibration consistency, the higher the similarity of each grain of rice in terms of vibration, and the more consistent the weight, specification and shape of each grain of rice, and the more stable the quality of the rice; A judging module, used to judge whether the vibration consistency is greater than or equal to a vibration consistency threshold; A transmitting module, configured to determine that the quality of the rice to be tested is unqualified if the vibration consistency is less than a vibration consistency threshold, and to control an ultrasonic transmitter to transmit an ultrasonic signal to the rice to be tested and receive an ultrasonic signal returned by the rice to be tested if the vibration consistency is greater than or equal to the vibration consistency threshold; A quality determination module, used for determining whether the quality of the rice to be detected is qualified by using an acoustic wave analysis model based on the ultrasonic signal returned by the rice to be detected; The system is also used to: Obtaining a heating video of the rice to be tested and electronic nose data during the heating process of the rice to be tested; Determine the cracking degree of the rice to be detected and the aroma degree of the rice to be detected using a heating data processing model based on the heating video of the rice to be detected and the electronic nose data during the heating process of the rice to be detected; If the cracking degree of the rice to be tested is greater than the cracking threshold, it is determined that the quality of the rice to be tested is unqualified; If the aroma degree of the rice to be detected is less than the aroma threshold, it is determined that the quality of the rice to be detected is unqualified.

5. The rice quality assessment system according to claim 4, characterized in that: The acoustic wave analysis model is an artificial neural network model, the input of the acoustic wave analysis model is the ultrasonic signal returned by the rice to be detected, and the output of the acoustic wave analysis model is whether the quality of the rice to be detected is qualified or unqualified.

6. The rice quality assessment system according to claim 4, characterized in that: The heating data processing model is a gated loop unit, the input of the heating data processing model is the heating video of the rice to be detected and the electronic nose data during the heating process of the rice to be detected, and the output of the heating data processing model is the cracking degree of the rice to be detected and the aroma degree of the rice to be detected.

Images