A method and device for detecting the quality of rubber accelerator particles based on perception fusion

Abstract

The application provides a rubber accelerator particle quality detection method and device based on perception fusion, and belongs to the field of artificial intelligence, and is used to improve the robustness of quality detection. The method comprises the following steps: an electronic device acquires images photographed for multiple rubber accelerator particles; the electronic device acquires a perception signal sequence obtained by perceiving the multiple rubber accelerator particles; the electronic device performs multi-modal dimension fusion detection processing on the images and the perception signal sequence through an AI / ML model to obtain a detection result, and the detection result indicates whether the multiple rubber accelerator particles include a rubber accelerator particle with a quality defect.

Description

Technical Field

[0001] This invention relates to the field of graph artificial intelligence technology, and in particular to a method and apparatus for detecting the quality of rubber accelerator particles based on perception fusion. Background Technology

[0002] Rubber accelerator granules are indispensable additives in the rubber industry. By regulating the vulcanization reaction rate and crosslinking density, they directly affect the mechanical properties (such as tensile strength and abrasion resistance) and processing efficiency of rubber products. In key areas such as tires, seals, and conveyor belts, the physical characteristics (particle size distribution, morphological uniformity, purity) and chemical characteristics (active ingredient content, dispersibility) of accelerator granules directly determine the quality and stability of the final product. Currently, with the development of neural network technology, convolutional neural networks (CNNs) have provided a breakthrough solution for intelligent quality inspection of rubber accelerator granules. For example, based on improved YOLOv8 or Mask R-CNN algorithms, real-time location and classification of defects such as particle agglomeration, foreign matter inclusions, and surface cracks can be achieved.

[0003] However, the robustness and reliability of CNN detection are not yet high enough. How to further improve the robustness and reliability of rubber accelerator particle quality detection is a current research problem. Summary of the Invention

[0004] This invention provides a method and apparatus for detecting the quality of rubber accelerator particles based on sensor fusion, in order to improve the robustness of quality detection.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] In a first aspect, a method for detecting the quality of rubber accelerator particles based on perceptual fusion is provided and applied to an electronic device. The method includes: the electronic device acquiring images of multiple rubber accelerator particles; the electronic device acquiring a perceptual signal sequence obtained by perceiving the multiple rubber accelerator particles; the electronic device performing multimodal dimensional fusion detection processing on the images and perceptual signal sequence using an AI / ML model to obtain a detection result, the detection result indicating whether there are rubber accelerator particles with quality defects among the multiple rubber accelerator particles.

[0007] Optionally, the electronic device acquires a sensing signal sequence obtained by sensing multiple rubber accelerator particles, including: the electronic device acquires the sensing signal sequence from a sensing receiver, wherein the sensing transmitter sends sensing signals to the multiple rubber accelerator particles, and the sensing receiver obtains the sensing signal sequence by receiving the echo signals of the sensing signals.

[0008] Optionally, multiple rubber accelerator particles are arranged in an array on the detection panel. The sensing transmitter sends K beams (K is an integer greater than 1) to the array in the direction pointing to the rows of the array. Any two beams of the K beams cover different rows of the array, and the K beams cover all rows of the array. Each of the K beams carries a sensing signal, resulting in a total of K sensing signals. The sensing receiver obtains a sensing signal sequence by receiving the echo signal of each of the K sensing signals, resulting in a total of K sensing signal sequences. Correspondingly, the electronic device acquires images of the multiple rubber accelerator particles, including: the electronic device acquiring images of the multiple rubber accelerator particles captured by the imaging device; wherein the imaging direction of the imaging device is pointed to and perpendicular to the detection panel, and the imaging device captures images of the multiple rubber accelerator particles to obtain images.

[0009] Optionally, the electronic device performs multimodal dimensional fusion detection processing on the image and the sensing signal sequence using an AI / ML model to obtain detection results, including: the electronic device extracts features from the image using the feature extraction layer of the AI / ML model to obtain image features; the electronic device divides the image features into K parts based on the number of sensing signal sequences being K, obtaining K sub-features, and fuses the K sub-features with the K sensing signal sequences one-to-one to obtain K fused features; the electronic device performs feature detection processing on the K fused features using the feature processing layer of the AI / ML model to obtain detection results.

[0010] Optionally, the image features are constructed as a matrix, with the sensing signal sequence corresponding to the elements in the rows of the matrix. Based on this, the electronic device divides the image features into K parts according to the number of sensing signal sequences (K), obtaining K sub-features, and fuses the K sub-features with the K sensing signal sequences one-to-one to obtain K fused features. This includes: the electronic device dividing the matrix along the column dimension according to the number of sensing signal sequences (K), thereby dividing the matrix into K sub-matrices. Each of the K sub-matrices is a sub-feature, and the column size of each of the K sub-matrices is floor(1 / K) of the column size of the matrix, where floor() represents rounding down; the electronic device fuses the K sub-matrices with the K sensing signal sequences one-to-one to obtain K fused features.

[0011] Optionally, the electronic device fuses the K subarrays with the K sensing signal sequences in a one-to-one correspondence to obtain K fused features, including: the electronic device redirects the element of the i-th sensing signal sequence in the K sensing signal sequences to the element of the i-th subarray in the K subarrays to obtain a fused feature, where i is an integer, and when i is an integer from 1 to K, a total of K fused features are obtained; wherein, the element of the i-th sensing signal sequence is redirected to the element of the i-th subarray, that is, the i-th sensing signal sequence is fused with the i-th subarray.

[0012] Optionally, each row of the array contains P rubber accelerator particles, where P is an integer greater than 1. The electronic device redirects the elements of the i-th sensing signal sequence in the K sensing signal sequences to the elements of the i-th subarray in the K subarrays to obtain a fused feature, including: the electronic device dividing the i-th sensing signal sequence into P sensing signal subsequences based on the fact that each row of the array contains P rubber accelerator particles; the electronic device fusing the elements contained in the j-th sensing signal subsequence in the P sensing signal subsequences into the S-th element of the i-th subarray. j To T j For each element in a column, a fused feature is obtained; the number of columns in the i-th subarray is M, where M is an integer greater than 1, and the S-th column... j To T j The number of columns is floor(M / P)*j; where the j-th sensing signal subsequence contains L elements, where L is an integer greater than 1. Fusion refers to combining the first S-th... j To T j In the construction of each element in the column, L zero elements are added, and then the L zero elements are replaced one-to-one with the elements contained in the j-th perceptual signal subsequence; or, fusion refers to combining the S-th... j To T j Each element of the column is added to the elements contained in the j-th sensing signal subsequence.

[0013] Optionally, each row of the array contains P rubber accelerator particles, where P is an integer greater than 1. The electronic device redirects the elements of the i-th sensing signal sequence in the K sensing signal sequences to the elements of the i-th subarray in the K subarrays to obtain a fused feature, including: the electronic device dividing the i-th sensing signal sequence into P sensing signal subsequences based on the fact that each row of the array contains P rubber accelerator particles; the electronic device randomly selects an element from the elements contained in the j-th sensing signal subsequence in the P sensing signal subsequences and fuses it into the S-th element of the i-th subarray. j To T j From the corresponding element in the column, a fused feature is obtained; the number of columns in the i-th subarray is M, where M is an integer greater than 1, and the S-th column... j To T jThe number of columns is floor(M / P)*j; where, fusion refers to randomly selecting one element from the elements contained in the j-th perceived signal subsequence and combining it with the S-th... j To T j Add the corresponding elements in the column.

[0014] Optionally, the sensing signal is a signal generated by a ZC sequence, an m sequence, or a gold sequence.

[0015] Secondly, a rubber accelerator particle quality detection device based on perceptual fusion is provided. The device is configured to: acquire images of multiple rubber accelerator particles; acquire a perceptual signal sequence obtained by perceiving the multiple rubber accelerator particles; and perform multimodal dimensional fusion detection processing on the images and perceptual signal sequence using an AI / ML model to obtain a detection result, the detection result indicating whether there are rubber accelerator particles with quality defects among the multiple rubber accelerator particles.

[0016] Optionally, acquiring a sensing signal sequence obtained by sensing multiple rubber accelerator particles includes: acquiring the sensing signal sequence from a sensing receiver, wherein a sensing transmitter sends sensing signals to multiple rubber accelerator particles, and the sensing receiver obtains the sensing signal sequence by receiving the echo signals of the sensing signals.

[0017] Optionally, multiple rubber accelerator particles are arranged in an array on the detection panel. The sensing transmitter sends K beams (K is an integer greater than 1) to the array in the direction pointing towards the rows of the array. Any two beams of the K beams cover different rows of the array, and the K beams cover all rows of the array. Each of the K beams carries a sensing signal, resulting in a total of K sensing signals. The sensing receiver obtains a sensing signal sequence by receiving the echo signal of each of the K sensing signals, resulting in a total of K sensing signal sequences. Correspondingly, images of the multiple rubber accelerator particles are acquired, including: acquiring images of the multiple rubber accelerator particles captured by the imaging device; wherein the imaging direction of the imaging device is pointed to and perpendicular to the detection panel, and the imaging device captures images of the multiple rubber accelerator particles to obtain images.

[0018] Optionally, a multimodal dimensional fusion detection process is performed on the image and the sensing signal sequence using an AI / ML model to obtain the detection result. This includes: performing feature extraction processing on the image using the feature extraction layer of the AI / ML model to obtain the image features; dividing the image features into K parts based on the number of sensing signal sequences (K), obtaining K sub-features, and fusing the K sub-features with the K sensing signal sequences one-to-one to obtain K fused features; and performing feature detection processing on the K fused features using the feature processing layer of the AI / ML model to obtain the detection result.

[0019] Optionally, the image features are constructed as a matrix, with the perceptual signal sequences corresponding to the elements in the rows of the matrix. Based on this, according to the number of perceptual signal sequences being K, the image features are divided into K parts to obtain K sub-features. The K sub-features are then fused with the K perceptual signal sequences in a one-to-one correspondence to obtain K fused features. This includes: dividing the matrix along the column dimension according to the number of perceptual signal sequences being K, thereby dividing the matrix into K sub-matrices. Each of the K sub-matrices is a sub-feature, and the column size of each of the K sub-matrices is floor(1 / K) of the column size of the matrix, where floor() represents rounding down; and fusing the K sub-matrices with the K perceptual signal sequences in a one-to-one correspondence to obtain K fused features.

[0020] Optionally, the K subarrays are fused one-to-one with the K sensing signal sequences to obtain K fused features, including: redirecting the element of the i-th sensing signal sequence in the K sensing signal sequences to the element of the i-th subarray in the K subarrays to obtain a fused feature, where i is an integer. When i is an integer from 1 to K, a total of K fused features are obtained; wherein, redirecting the element of the i-th sensing signal sequence to the element of the i-th subarray means that the i-th sensing signal sequence is fused with the i-th subarray.

[0021] Optionally, each row of the array contains P rubber accelerator particles, where P is an integer greater than 1. Elements of the i-th sensing signal sequence in the K sensing signal sequences are redirected to elements of the i-th subarray in the K subarrays to obtain a fused feature. This includes: dividing the i-th sensing signal sequence into P sensing signal subsequences based on the P rubber accelerator particles in each row of the array; and fusing the elements of the j-th sensing signal subsequence in the P sensing signal subsequences into the S-th element of the i-th subarray. j To T j For each element in a column, a fused feature is obtained; the number of columns in the i-th subarray is M, where M is an integer greater than 1, and the S-th column... j To T j The number of columns is floor(M / P)*j; where the j-th sensing signal subsequence contains L elements, where L is an integer greater than 1. Fusion refers to combining the first S-th... j To T j In the construction of each element in the column, L zero elements are added, and then the L zero elements are replaced one-to-one with the elements contained in the j-th perceptual signal subsequence; or, fusion refers to combining the S-th... j To T j Each element of the column is added to the elements contained in the j-th sensing signal subsequence.

[0022] Optionally, each row of the array contains P rubber accelerator particles, where P is an integer greater than 1. The elements of the i-th sensing signal sequence in the K sensing signal sequences are redirected to the elements of the i-th subarray in the K subarrays to obtain a fused feature. This includes: dividing the i-th sensing signal sequence into P sensing signal subsequences based on the P rubber accelerator particles in each row of the array; randomly selecting an element from the elements of the j-th sensing signal subsequence in the P sensing signal subsequences and fusing it into the S-th element of the i-th subarray. j To T j From the corresponding element in the column, a fused feature is obtained; the number of columns in the i-th subarray is M, where M is an integer greater than 1, and the S-th column... j To T j The number of columns is floor(M / P)*j; where, fusion refers to randomly selecting one element from the elements contained in the j-th perceived signal subsequence and combining it with the S-th... j To T j Add the corresponding elements in the column.

[0023] Optionally, the sensing signal is a signal generated by a ZC sequence, an m sequence, or a gold sequence.

[0024] Thirdly, a computer-readable storage medium is provided, comprising: a computer program or instructions; when the computer program or instructions are executed on a computer, the computer causes the computer to perform the method described in the first aspect.

[0025] In summary, the above method and apparatus have the following technical effects:

[0026] Electronic devices can acquire images of multiple rubber accelerator particles and a sensing signal sequence obtained by sensing the multiple rubber accelerator particles. Then, the electronic devices use an AI / ML model to perform multimodal dimensional fusion detection processing on the images and sensing signal sequence to obtain detection results of rubber accelerator particles indicating whether there are quality defects in the multiple rubber accelerator particles. By fusing perception and image processing, the robustness and reliability of rubber accelerator particle quality detection can be further improved. Attached Figure Description

[0027] Figure 1 A schematic flowchart of the rubber accelerator particle quality detection method based on sensor fusion provided in an embodiment of the present invention;

[0028] Figure 2 A schematic diagram of a scenario for the rubber accelerator particle quality detection method based on sensor fusion provided in an embodiment of the present invention;

[0029] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0030] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0031] This invention will be presented in relation to systems that may include multiple devices, components, modules, etc. It should be understood and appreciated that various systems may include additional devices, components, modules, etc., and / or may not include all the devices, components, modules, etc. discussed in conjunction with the accompanying drawings. Furthermore, combinations of these approaches are also possible.

[0032] Furthermore, in embodiments of the present invention, words such as "exemplary" and "for example" are used to indicate that something is presented as an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Rather, the use of the term "exemplary" is intended to present the concept in a specific manner.

[0033] In the embodiments of this invention, "of", "corresponding (relevant)" and "corresponding" can sometimes be used interchangeably. It should be noted that when their distinction is not emphasized, their intended meanings are matching. In addition, the " / " mentioned in this invention can be used to indicate an "or" relationship.

[0034] The network architecture and business scenarios described in the embodiments of this invention are for the purpose of more clearly illustrating the technical solutions of the embodiments of this invention, and do not constitute a limitation on the technical solutions provided by the embodiments of this invention. As those skilled in the art will know, with the evolution of network architecture and the emergence of new business scenarios, the technical solutions provided by the embodiments of this invention are also applicable to similar technical problems.

[0035] For example, Figure 1 This invention provides a schematic flowchart of a method for detecting the quality of rubber accelerator particles based on sensor fusion. This method can be applied to electronic devices.

[0036] like Figure 1 As shown, the flowchart of this rubber accelerator particle quality detection method based on sensor fusion is as follows:

[0037] S101, the electronic device acquires images of multiple rubber accelerator particles.

[0038] S102, the electronic device acquires a sequence of sensing signals obtained by sensing multiple rubber accelerator particles.

[0039] In steps S101-S102, the electronic device can acquire a sensing signal sequence from the sensing receiver. Specifically, the sensing transmitter sends sensing signals to multiple rubber accelerator particles, and the sensing receiver obtains the sensing signal sequence by receiving the echo signals of these sensing signals. The multiple rubber accelerator particles are arranged in an array on the detection panel. The sensing transmitter sends K beams (where K is an integer greater than 1) to the array in the direction pointing towards the rows of the array. Any two beams from the K beams cover different rows of the array, and the K beams cover all rows of the array. Each of the K beams carries one sensing signal, resulting in a total of K sensing signals. These sensing signals are generated using a ZC sequence, an m-sequence, or a gold sequence; that is, the sensing transmitter can generate and transmit sensing signals using a ZC sequence, an m-sequence, or a gold sequence. The sensing receiver obtains a sensing signal sequence by receiving the echo signals of each of the K sensing signals, resulting in a total of K sensing signal sequences. Correspondingly, the electronic device can acquire images of multiple rubber accelerator particles captured by the imaging device; wherein, the imaging direction of the imaging device is pointed to and perpendicular to the detection panel, and the imaging device can capture images of multiple rubber accelerator particles to obtain images.

[0040] In this context, each rubber accelerator particle in a plurality of rubber accelerator particles may be agglomerated.

[0041] For ease of understanding, such as Figure 2 As shown, multiple rubber accelerator particles are placed in a 6x6 array on the detection panel. The sensing transmitter transmits three beams along the rows of the array: beam #1, beam #2, and beam #3. Beam #1 covers rows 1-2 of the array, totaling 12 rubber accelerator particles. Beam #2 covers rows 3-4, totaling 12 rubber accelerator particles. Beam #3 covers rows 5-6, totaling 12 rubber accelerator particles. The sensing receiver receives the echoes generated by the reflection / refraction of the three beams by the rubber accelerator particles, thus receiving three sensing signal sequences.

[0042] S103, the electronic device uses an AI / ML model to perform multimodal dimensional fusion detection processing on the image and the perceived signal sequence to obtain the detection result.

[0043] The test results indicate whether there are quality defects in the rubber accelerator particles among the multiple rubber accelerator particles.

[0044] The following is a detailed introduction to S103:

[0045] First, electronic devices can extract features from images using the feature extraction layer of an artificial intelligence (AI) / machine learning (ML) model. The AI / ML model can be a deep neural network model, such as a convolutional neural network (CNN). The feature extraction layer can perform convolution processing, extracting image features through convolution. The image features are constructed as a matrix, such as an M*N matrix, where M and N are both integers greater than 1, M is the number of columns, and N is the number of rows.

[0046] Secondly, the electronic device can divide the features of the image into K parts based on the number of sensing signal sequences, obtain K sub-features, and fuse the K sub-features with the K sensing signal sequences one-to-one to obtain K fused features.

[0047] For example, the sensing signal sequence corresponds to the elements in the rows of a matrix. That is, for each sensing signal sequence, the elements in the sequence, ordered from front to back (or from smallest to largest index), can sequentially characterize the features of one or more rows of rubber accelerator particles in the corresponding array. Similarly, the elements in a row of a matrix, ordered from front to back (or from smallest to largest index), can sequentially characterize the features of one or more rows of rubber accelerator particles in the corresponding array. Based on this, the electronic device can divide the matrix into K subarrays along the column dimension, based on the number of K sensing signal sequences. Each subarray represents a sub-feature, and the column size of each subarray is floor(1 / K) of the matrix's column size, where floor() represents rounding down. For example, a 7x6 matrix, where each row has 6 elements (each element is a vector) and each column has 7 elements, such as... Figure 2 In the scenario shown, K=3, meaning the 7x6 matrix can be divided into 3 subarrays. The first subarray is a 2x7 subarray, where the 2 rows are the first and 2nd rows of the matrix, and the size of each column (or the number of elements in each column) is 2. The second subarray is a 2x7 subarray, where the 2 rows are the third and 4th rows of the matrix, and the size of each column is also 2. The third subarray is a 2 / 3x7 subarray, where the 2 / 3 rows are the fifth to sixth / seventh rows of the matrix, and the size of each column can be 2 or 3.

[0048] Electronic devices can fuse K subarrays with K sensing signal sequences in a one-to-one correspondence to obtain K fused features. For example, an electronic device can redirect the element of the i-th sensing signal sequence in the K sensing signal sequences to the element of the i-th subarray in the K subarrays to obtain a fused feature, where i is an integer. If i is an integer ranging from 1 to K, a total of K fused features are obtained. Redirecting the element of the i-th sensing signal sequence to the element of the i-th subarray means fusing the i-th sensing signal sequence with the i-th subarray. This redirection can be understood as matching the physical spatial positions during fusion to achieve the fusion of elements of the same rubber accelerator particle in the image with elements in the sensing signal sequence.

[0049] In one possible approach, each row of the array can contain P rubber accelerator particles, where P is an integer greater than 1. The electronic device can divide the i-th sensing signal sequence into P sensing signal sub-sequences based on the P rubber accelerator particles in each row of the array. The electronic device can then fuse the elements contained in the j-th sensing signal sub-sequence among the P sensing signal sub-sequences into the S-th element of the i-th subarray. j To T j For each element in a column, a fused feature is obtained; the number of columns in the i-th subarray is M, where M is an integer greater than 1, and the S-th column... j To T j The number of columns is floor(M / P)*j.

[0050] Here, the j-th sensing signal subsequence contains L elements, where L is an integer greater than 1. Fusion refers to combining the first S-th sensing signal subsequence with the second S-th sensing signal subsequence. j To T j In the construction of each element in the column, L zero elements are added, and then the L zero elements are replaced one-to-one with the elements contained in the j-th perceptual signal subsequence; or, fusion refers to combining the S-th... j To T j Each element of the column is added to the elements contained in the j-th sensing signal subsequence.

[0051] For example, S of the i-th subarray j To T j The following formula (1) is used:

[0052]

[0053] The j-th sensing signal subsequence is shown in equation (2) below:

[0054] (y1, y2, y3); (2)

[0055] Based on this, the characteristics of fusion can be shown in equation (3) / (4);

[0056]

[0057] Here, x11 represents one element, x12 represents another element, and so on. Similarly, y1 also represents one element, y2 represents another element, and so on.

[0058] In another possible approach, each row of the array contains P rubber accelerator particles, where P is an integer greater than 1. The electronic device can divide the i-th sensing signal sequence into P sensing signal sub-sequences based on the P rubber accelerator particles per row. The electronic device then randomly selects an element from the elements of the j-th sensing signal sub-sequence among the P sensing signal sub-sequences and merges it into the S-th subarray of the i-th subarray. j To T j From the corresponding element in the column, a fused feature is obtained; the number of columns in the i-th subarray is M, where M is an integer greater than 1, and the S-th column... j To T j The number of columns is floor(M / P)*j. Here, fusion refers to randomly selecting one element from the elements contained in the j-th perceived signal subsequence and combining it with the S-th... j To T j Add the corresponding elements in the column.

[0059] It is understandable that the above random sampling method involves selecting the Sth... j To T j The columns are considered as a whole, which, after fusion, can contain elements of the corresponding sensing signal subsequences.

[0060] Finally, the electronic device can perform feature detection processing on the K fused features through the feature processing layer of the AI / ML model to obtain the detection result. The feature processing layer can include a pooling layer, a fully connected layer, and an output layer. The detection result can be a rubber accelerator particle marked with a quality defect in the image, indicating that the rubber accelerator particle has a quality defect, or the image itself, indicating that the rubber accelerator particle does not have a quality defect.

[0061] In summary, electronic devices can acquire images of multiple rubber accelerator particles and a sequence of sensing signals obtained by sensing these particles. Then, using an AI / ML model, the electronic device performs multimodal dimensional fusion detection processing on the images and the sensing signal sequence to obtain detection results indicating whether quality defects exist in the multiple rubber accelerator particles. By fusing sensing and image processing, the robustness and reliability of rubber accelerator particle quality detection can be further improved.

[0062] The above combination Figure 1This invention provides a detailed description of a method for detecting the quality of rubber accelerator particles based on sensor fusion, as provided in the embodiments of the present invention. The following details a sensor fusion-based apparatus for detecting the quality of rubber accelerator particles used to perform the method provided in the embodiments of the present invention.

[0063] The device is configured to: acquire images of multiple rubber accelerator particles; acquire a sequence of sensing signals obtained by sensing the multiple rubber accelerator particles; and perform multimodal dimensional fusion detection processing on the images and sensing signal sequences using an AI / ML model to obtain detection results, which indicate whether there are rubber accelerator particles with quality defects among the multiple rubber accelerator particles.

[0064] Optionally, acquiring a sensing signal sequence obtained by sensing multiple rubber accelerator particles includes: acquiring the sensing signal sequence from a sensing receiver, wherein a sensing transmitter sends sensing signals to multiple rubber accelerator particles, and the sensing receiver obtains the sensing signal sequence by receiving the echo signals of the sensing signals.

[0065] Optionally, multiple rubber accelerator particles are arranged in an array on the detection panel. The sensing transmitter sends K beams (K is an integer greater than 1) to the array in the direction pointing towards the rows of the array. Any two beams of the K beams cover different rows of the array, and the K beams cover all rows of the array. Each of the K beams carries a sensing signal, resulting in a total of K sensing signals. The sensing receiver obtains a sensing signal sequence by receiving the echo signal of each of the K sensing signals, resulting in a total of K sensing signal sequences. Correspondingly, images of the multiple rubber accelerator particles are acquired, including: acquiring images of the multiple rubber accelerator particles captured by the imaging device; wherein the imaging direction of the imaging device is pointed to and perpendicular to the detection panel, and the imaging device captures images of the multiple rubber accelerator particles to obtain images.

[0066] Optionally, a multimodal dimensional fusion detection process is performed on the image and the sensing signal sequence using an AI / ML model to obtain the detection result. This includes: performing feature extraction processing on the image using the feature extraction layer of the AI / ML model to obtain the image features; dividing the image features into K parts based on the number of sensing signal sequences (K), obtaining K sub-features, and fusing the K sub-features with the K sensing signal sequences one-to-one to obtain K fused features; and performing feature detection processing on the K fused features using the feature processing layer of the AI / ML model to obtain the detection result.

[0067] Optionally, the image features are constructed as a matrix, with the perceptual signal sequence corresponding to the elements in the rows of the matrix. Based on this, according to the number of perceptual signal sequences being K, the image features are divided into K parts to obtain K sub-features. The K sub-features are then fused with the K perceptual signal sequences in a one-to-one correspondence to obtain K fused features. This includes: dividing the matrix along the column dimension according to the number of perceptual signal sequences being K, thereby dividing the matrix into K sub-matrices. Each of the K sub-matrices is a sub-feature. The column size of each of the K sub-matrices is floor(1 / K) of the column size of the matrix, where floor() represents rounding down. The K sub-matrices are then fused with the K perceptual signal sequences in a one-to-one correspondence to obtain K fused features.

[0068] Optionally, the K subarrays are fused one-to-one with the K sensing signal sequences to obtain K fused features, including: redirecting the element of the i-th sensing signal sequence in the K sensing signal sequences to the element of the i-th subarray in the K subarrays to obtain a fused feature, where i is an integer. When i is an integer from 1 to K, a total of K fused features are obtained; wherein, redirecting the element of the i-th sensing signal sequence to the element of the i-th subarray means that the i-th sensing signal sequence is fused with the i-th subarray.

[0069] Optionally, each row of the array contains P rubber accelerator particles, where P is an integer greater than 1. Elements of the i-th sensing signal sequence in the K sensing signal sequences are redirected to elements of the i-th subarray in the K subarrays to obtain a fused feature. This includes: dividing the i-th sensing signal sequence into P sensing signal subsequences based on the P rubber accelerator particles in each row of the array; and fusing the elements of the j-th sensing signal subsequence in the P sensing signal subsequences into the S-th element of the i-th subarray. j To T j For each element in a column, a fused feature is obtained; the number of columns in the i-th subarray is M, where M is an integer greater than 1, and the S-th column... j To T j The number of columns is floor(M / P)*j; where the j-th sensing signal subsequence contains L elements, where L is an integer greater than 1. Fusion refers to combining the first S-th... j To T j In the construction of each element in the column, L zero elements are added, and then the L zero elements are replaced one-to-one with the elements contained in the j-th perceptual signal subsequence; or, fusion refers to combining the S-th... j To T j Each element of the column is added to the elements contained in the j-th sensing signal subsequence.

[0070] Optionally, each row of the array contains P rubber accelerator particles, where P is an integer greater than 1. The elements of the i-th sensing signal sequence in the K sensing signal sequences are redirected to the elements of the i-th subarray in the K subarrays to obtain a fused feature. This includes: dividing the i-th sensing signal sequence into P sensing signal subsequences based on the P rubber accelerator particles in each row of the array; randomly selecting an element from the elements of the j-th sensing signal subsequence in the P sensing signal subsequences and fusing it into the S-th element of the i-th subarray. j To T j From the corresponding element in the column, a fused feature is obtained; the number of columns in the i-th subarray is M, where M is an integer greater than 1, and the S-th column... j To T j The number of columns is floor(M / P)*j; where, fusion refers to randomly selecting one element from the elements contained in the j-th perceived signal subsequence and combining it with the S-th... j To T j Add the corresponding elements in the column.

[0071] Optionally, the sensing signal is a signal generated by a ZC sequence, an m sequence, or a gold sequence.

[0072] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Exemplarily, the electronic device may be a terminal device, or a chip (system) or other component or assembly that can be disposed in the terminal device. Figure 3 As shown, the electronic device 400 may include a processor 401. Optionally, the electronic device 400 may also include a memory 402 and / or a transceiver 403. The processor 401 is coupled to the memory 402 and the transceiver 403, for example, they can be connected via a communication bus. Alternatively, the electronic device 400 may also be a chip, such as including the processor 401; in this case, the transceiver may be the chip's input / output interface.

[0073] The following is combined with Figure 3 The various components of electronic device 400 are described in detail below:

[0074] The processor 401 is the control center of the electronic device 400. It can be a single processor or a collective term for multiple processing elements. For example, the processor 401 can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement embodiments of the present invention, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0075] Optionally, the processor 401 can perform various functions of the electronic device 400, such as the aforementioned functions, by running or executing software programs stored in the memory 402 and calling scientific data stored in the memory 402. Figure 1 The method for detecting the quality of rubber accelerator particles based on sensor fusion is shown.

[0076] In a specific implementation, as one example, processor 401 may include one or more CPUs, for example... Figure 3 CPU0 and CPU1 are shown in the diagram.

[0077] In a specific implementation, as one example, the electronic device 400 may also include multiple processors. Each of these processors may be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process scientific data (e.g., computer programs or instructions).

[0078] The memory 402 is used to store the software program that executes the solution of the present invention, and is controlled by the processor 401 to execute it. The specific implementation method can be referred to the above method embodiment, and will not be repeated here.

[0079] Optionally, the memory 402 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or scientific data structures and accessible by a computer, but not limited thereto. The memory 402 may be integrated with the processor 401 or may exist independently and be accessible through the interface circuit of the electronic device 400. Figure 3 (Not shown in the image) is coupled to processor 401, and this embodiment of the invention does not specifically limit this.

[0080] Transceiver 403 is used for communication with other electronic devices. For example, if electronic device 400 is a terminal device, transceiver 403 can be used to communicate with a network device or with another terminal device. As another example, if electronic device 400 is a network device, transceiver 403 can be used to communicate with a terminal device or with another network device.

[0081] Alternatively, transceiver 403 may include a receiver and a transmitter. Figure 3 (Not shown separately). The receiver is used to implement the receiving function, and the transmitter is used to implement the sending function.

[0082] Alternatively, the transceiver 403 can be integrated with the processor 401, or it can exist independently and be connected via the interface circuit of the electronic device 400. Figure 3 (Not shown in the image) is coupled to processor 401, and this embodiment of the invention does not specifically limit this.

[0083] Understandable Figure 3 The structure of the electronic device 400 shown does not constitute a limitation on the electronic device. Actual electronic devices may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0084] Furthermore, the technical effects of the electronic device 400 can be referred to the technical effects of the methods described in the above method embodiments, and will not be repeated here.

[0085] It should be understood that the processor in the embodiments of the present invention can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0086] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer program or instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another.

[0087] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0088] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0089] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0090] In the embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0091] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0092] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for detecting the quality of rubber accelerator particles based on perception fusion, characterized in that, The method is applied to an electronic device, and the method comprises: The electronic device obtains an image of a plurality of rubber accelerator particles; The electronic device obtains a perception signal sequence obtained by perceiving the plurality of rubber accelerator particles; The electronic device performs multi-modal dimension fusion detection processing on the image and the perception signal sequence by an AI / ML model to obtain a detection result, and the detection result indicates whether the plurality of rubber accelerator particles include a rubber accelerator particle with a quality defect; The electronic device obtains a perception signal sequence obtained by perceiving the plurality of rubber accelerator particles, comprising: The electronic device obtains the perception signal sequence from a perception receiver, wherein a perception transmitter sends a perception signal to the plurality of rubber accelerator particles, and the perception receiver obtains the perception signal sequence by receiving echo signals of the perception signal; The plurality of rubber accelerator particles are arranged in an array on a detection panel, the perception transmitter sends K beams in a direction pointing to rows of the array, K is an integer greater than 1, any two beams of the K beams cover different rows of the array, the K beams cover all rows of the array, each beam of the K beams corresponds to a perception signal, there are K perception signals, the perception receiver obtains a perception signal sequence by receiving echo signals of each perception signal of the K perception signals, and there are K perception signal sequences; Correspondingly, the electronic device obtains an image of a plurality of rubber accelerator particles, comprising: The electronic device obtains the image from a shooting device, wherein a shooting direction of the shooting device points to and is perpendicular to the detection panel, the shooting device shoots the plurality of rubber accelerator particles to obtain the image; The electronic device performs multi-modal dimension fusion detection processing on the image and the perception signal sequence by an AI / ML model to obtain a detection result, comprising: The electronic device performs feature extraction processing on the image by a feature extraction layer of the AI / ML model to obtain a feature of the image; The electronic device divides the feature of the image into K parts according to the number K of the perception signal sequences, obtains K sub-features, and fuses the K sub-features with the K perception signal sequences one by one to obtain K fused features; The electronic device performs feature detection processing on the K fused features by a feature processing layer of the AI / ML model to obtain the detection result; The feature of the image is a matrix, and the perception signal sequence corresponds to an element in a row of the matrix; on this basis, the electronic device divides the feature of the image into K parts according to the number K of the perception signal sequences, obtains K sub-features, and fuses the K sub-features with the K perception signal sequences one by one to obtain K fused features, comprising: The electronic device divides the matrix into K sub-matrices in the dimension of columns of the matrix, so as to divide the matrix into K sub-features, and the size of the column of each of the K sub-matrices is floor(1 / K) of the size of the column of the matrix, where floor() represents rounding down. The electronic device fuses the K sub-matrices with the K perception signal sequences one by one to obtain the K fused features. The electronic device fuses the K sub-matrices with the K perception signal sequences one by one to obtain the K fused features, including: The electronic device redirects elements of an i-th perception signal sequence in the K perception signal sequences to elements of an i-th sub-matrix in the K sub-matrices to obtain a fused feature, where i is an integer, and the K fused features are obtained when i is an integer traversing from 1 to K; and the elements of the i-th perception signal sequence are redirected to the elements of the i-th sub-matrix, that is, the i-th perception signal sequence is fused with the i-th sub-matrix. Each row in the array contains P rubber accelerator particles, and P is an integer greater than 1. The electronic device redirects elements of an i-th perception signal sequence in the K perception signal sequences to elements of an i-th sub-matrix in the K sub-matrices to obtain a fused feature, including: The electronic device divides the i-th perception signal sequence into P perception signal subsequences according to that each row in the array contains the P rubber accelerator particles. The electronic device fuses the elements contained in the j-th sensing signal subsequence among the P sensing signal subsequences into the S-th element of the i-th subarray. j To T j In each element of the column, a fused feature is obtained; the number of columns of the i-th subarray is M, where M is an integer greater than 1, and the S-th... j To T j The number of columns is floor(M / P)*j; Wherein, the number of elements contained in the jth sensing signal subsequence is L, L is an integer greater than 1, and fusion refers to replacing each element in the first S j to T j column with L 0 elements, and then replacing the L 0 elements one by one with the elements contained in the jth sensing signal subsequence; or, fusion refers to adding each element in the first S j to T j column to the elements contained in the jth sensing signal subsequence.

2. The method of claim 1, wherein, Each row in the array contains P rubber accelerator particles, and P is an integer greater than 1. The electronic device redirects elements of an i-th perception signal sequence in the K perception signal sequences to elements of an i-th sub-matrix in the K sub-matrices to obtain a fused feature, including: The electronic device divides the i-th perception signal sequence into P perception signal subsequences according to that each row in the array contains the P rubber accelerator particles. The electronic device fuses one element randomly selected from elements contained in the jth perception signal subsequence of the P perception signal subsequences into a corresponding element in the Sth column of the ith submatrix, to obtain a fused feature; the number M of columns of the ith submatrix is an integer greater than 1, and the number of columns of the Sth submatrix is floor(M / P)*j. j to T j j to T j ​​ Wherein, the fusion refers to adding one element randomly selected from elements contained in the jth perception signal subsequence and one element corresponding to the Sth column in the Tth row. j to T j column.

3. The method according to claim 1 or 2, characterized in that, The perception signal is a signal generated by a ZC sequence, an m sequence, or a gold sequence.

Images