An information identification and processing system for solid particulate matter collection and testing

Abstract

This invention discloses an information identification and processing system for solid particulate matter collection and testing, belonging to the field of information identification and processing technology. It addresses the problems of existing solid particulate matter identification and processing technologies, which lack multi-angle identification and processing of solid particulate matter and lack retrospective correction of images used for solid particulate matter identification in the database, resulting in incomplete and inaccurate identification and detection results. The system includes a server and multiple detection terminals connected to the server. It processes received particulate matter information to obtain color mean, crack mean, incompleteness coefficient, and grayscale difference, normalizes these values, and analyzes them to obtain standard values. Based on the standard values, it compares and analyzes them with set corresponding thresholds, outputs and displays the detection results, and achieves multi-angle detection of particulate matter for a more comprehensive and complete detection. Simultaneously, it performs retrospective correction operations on the detection results to improve the accuracy of the detection results.

Description

Technical Field

[0001] This invention relates to the field of information identification and processing technology, specifically to an information identification and processing system for the collection and inspection of solid particulate matter. Background Technology

[0002] Currently, the identification and classification of solid particulate matter is mainly achieved through image-based rapid solid particulate matter detectors, aiming to improve the quality of the corresponding solid particulate matter. However, existing solid particulate matter identification and processing technologies have the following drawbacks: 1. Image-based identification and classification of solid particulate matter results in a limited range of methods, leading to incomplete detection results; 2. Environmental differences and instrument errors can cause erroneous detection results, resulting in inaccurate assessments. Summary of the Invention

[0003] The purpose of this invention is to address the problems of existing solid particulate matter identification and processing technologies, which lack multi-angle identification and processing of solid particulate matter and lack retrospective correction of photos used for solid particulate matter identification in the database, resulting in incomplete and inaccurate identification and detection results. Therefore, this invention proposes an information identification and processing system for solid particulate matter collection and inspection.

[0004] The objective of this invention can be achieved through the following technical solution: an information identification and processing system for collecting and testing solid particulate matter, comprising a server and multiple detection terminals communicatively connected to the server;

[0005] The detection end processes the received particulate matter information into color mean, crack mean, defect coefficient, and grayscale difference, normalizes them, takes their values, analyzes the values ​​to obtain standard quality values, compares and analyzes them with the set corresponding thresholds, outputs the detection results and displays the explanation; the detection end increments the detection count by one, and sends the detection results and the corresponding particulate matter information to the server;

[0006] The server establishes a database based on the detection results. The database stores all particulate matter parameters, images, and corresponding detection results uploaded by all detection terminals. Based on the detection results, the corresponding particulate matter parameters and images are classified and stored in different databases. A backtracking correction operation is performed on the particulate matter information in all databases to obtain correction results. The correction results include incorrect and correct results. When the correction result is correct, no operation is performed. When the correction result is incorrect, the particulate matter information is corrected and adjusted, and a new detection result is output. The new detection result is sent to the detection terminal to overwrite the original detection result, and the particulate matter information and corresponding detection results in all databases are updated and saved.

[0007] In a preferred embodiment of the present invention, the specific steps for generating the color mean value are as follows:

[0008] The system acquires the color and brightness values ​​of particulate matter, identifies the particulate matter, retrieves the corresponding standard color and brightness values ​​from the database, and combines the particulate matter's color value (ys), brightness value (lt), standard color value (bs), and standard brightness value (bt) using a preset formula. The color vividness value Sz is obtained, where a1, a2 and a3 are preset weighting coefficients;

[0009] The average color vividness value is obtained by retrieving the color vividness values ​​of all particulate matter in the test sample and taking the average value. It will be achieved through a preset formula. The color mean value SJZ is obtained, where c1 and c2 are preset weight coefficients.

[0010] In a preferred embodiment of the present invention, the specific steps for generating the split mean value are as follows:

[0011] Acquire particulate matter images, divide them into several grids, and count the number and area of ​​each grid. When a crack is detected in a grid, mark it as a crack and calculate its area. Sum all the crack areas to obtain the total crack area. Then, combine the grid area s, the number of cracks nu, and the total crack area sm using a preset formula. The apparent value BGZ is obtained; where b1 and b2 are preset weighting factors, and β is a preset correction coefficient.

[0012] Obtain the apparent values ​​of all particulate matter in the test sample, and select the median apparent value and the mode apparent value; calculate the median apparent value ZSB and the mode apparent value BGZj, where j = 1, 2, 3…n3, i ∈ j and n3 ∈ n2; and apply a preset formula... The mean value of the crack is obtained as LJZ, where d1 and d2 are preset weighting coefficients.

[0013] In a preferred embodiment of the present invention, the specific steps for generating the incompleteness coefficient are as follows:

[0014] Acquire particulate matter images, identify the edge positions of the particles, construct a two-dimensional coordinate system, and connect the leftmost and rightmost edge coordinates to segment the particulate matter 2D image into upper and lower curved sections. Divide the image into m regions of equal length, obtain the height of each point within each region, and calculate the average height by averaging the heights of all points within each region. Then, combine the height hr with the average height. By preset formula The gap value QKZ is obtained, where r = 1, 2, 3...n4, and n4 is a positive integer;

[0015] The total gap value is obtained by summing the gap values ​​in all regions of particulate matter. The total gap value is then compared with the corresponding set thresholds. Particulate matter is labeled as severely deficient, moderately deficient, and lightly deficient, and the corresponding quantities are counted and labeled as mu1, mu2, and mu3, respectively. When mu1 ≥ mu2 + mu3, the severe deficiency coefficient is matched; when mu3 ≥ mu1 + mu2, the light deficiency coefficient is matched; otherwise, the moderate deficiency coefficient is matched. The severe deficiency coefficient, moderate deficiency coefficient, and light deficiency coefficient are labeled as the residual coefficient Qp, where p = 1, 2, 3, p1 represents the severe deficiency coefficient, p2 represents the moderate deficiency coefficient, and p3 represents the light deficiency coefficient.

[0016] In a preferred embodiment of the present invention, the specific steps for generating grayscale difference are as follows:

[0017] The grayscale value inside the particulate matter is obtained by performing spectral analysis. A standard grayscale value is preset for each different particulate matter. The difference between the grayscale value and the marked grayscale value is calculated to obtain the grayscale difference value.

[0018] In a preferred embodiment of the present invention, the specific steps for the backtracking unit to perform backtracking traversal correction on the detection results are as follows:

[0019] S1: Establish a first database and a second database based on the particulate matter detection results, and save the detection results and the corresponding particulate matter information to the first database and the second database respectively;

[0020] S2: Select any detection result from the first database, select the target detection end, and send the corresponding particulate matter information to the target detection end for detection and judgment to obtain the correction result. When the received correction result is incorrect, the particulate matter information is corrected and adjusted, and the detection result is output again. The detection end is recorded as passively correcting once, the target detection end is recorded as actively correcting once, and the corresponding detection result is recorded as the corrected detection result. When the received correction result is correct, the first particulate matter sample prompt is output, and particulate matter is selected for detection and judgment again.

[0021] In a preferred embodiment of the present invention, the specific steps for selecting the target detection end are as follows:

[0022] Obtain the detection terminal corresponding to the detection result, and use a preset formula to calculate the number of detections T1, the number of active error corrections T2, and the number of passive error corrections T3 of other detection terminals. The judgment value TPZ is obtained, where z1, z2 and z are preset weight coefficients, and μ is a preset correction coefficient. The detection end with the largest judgment value is recorded as the target detection end. The target detection end receives particulate matter information, performs particulate matter detection, and outputs the detection result again.

[0023] Compared with the prior art, the beneficial effects of the present invention are:

[0024] 1. The received particulate matter information is processed to obtain color mean, crack mean, defect coefficient and gray difference value. The normalized values ​​are taken and the values ​​are analyzed to obtain standard values. The standard values ​​are compared with the set corresponding thresholds to output the detection results and display the explanation. This realizes the detection of particulate matter from multiple perspectives, and the detection of particulate matter is more comprehensive and complete.

[0025] 2. By classifying and backtracking the test results, a suitable testing end is selected for retesting. When the retest result matches the original test result, any test result is randomly selected for verification, ensuring that all test results are traversed. When the test results are inconsistent, the retest result is output and overwritten. The testing end with the result is recorded as having undergone passive error correction once. The number of passive error corrections for each testing end is counted. When the number of passive error corrections exceeds a preset number, the testing end is taken offline and notified for maintenance. This not only enables backtracking and correction of test results, improving their accuracy, but also allows for the offline shutdown of testing ends with more than a preset number of errors, thus monitoring the testing end and improving the accuracy of the test results. Attached Figure Description

[0026] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0027] Figure 1 This is the overall system block diagram of the present invention;

[0028] Figure 2 This is a particulate edge diagram of the present invention. Detailed Implementation

[0029] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] Please see Figure 1 As shown, an information identification and processing system for solid particulate matter collection and testing includes a server and multiple detection terminals wirelessly connected to the server; the solid particulate matter to be detected is sent to the detection terminal, and the detection terminal obtains particulate matter information through multiple sensors and camera modules, wherein the particulate matter information includes particulate matter parameters and images, the particulate matter parameters are color values ​​and brightness values, and the images are internal spectrum diagrams and particulate matter images;

[0031] The detection device identifies particulate matter information and outputs detection results. The specific steps are as follows:

[0032] Each particulate matter is pre-defined to have a standard color value and a standard brightness value. The current particulate matter is matched with all other particulate matter to obtain its corresponding standard color value and standard brightness value. The color value and brightness value of the particulate matter are then obtained and labeled as ys and lt, respectively. These values ​​are then compared with the standard color value bs and the standard brightness value bt using a pre-defined formula. The color vividness value Sz is obtained, where a1, a2, and a3 are preset weighting coefficients. The formula shows that the smaller the absolute value of the color value, brightness value, and standard color value, the smaller the difference between the current particulate matter and the standard particulate matter; the larger the color vividness value, the better the quality of the particulate matter. Each color vividness value corresponds to one preset value. It should be noted that the color value represents the degree of color intensity; the darker the color, the larger the color value, and the lighter the color, the smaller the color value. Similarly, the brightness value represents the degree of brightness.

[0033] If we take i particles as a test sample, where i = 1, 2, 3...n2, and n2 is a positive integer; obtain the color vividness value Szi corresponding to all particles in the test sample and take the average value to get the average color vividness value. Through preset formula The color mean value SJZ is obtained, where c1 and c2 are preset weighting coefficients. The lower the dispersion of the color vividness value, the larger the color vividness value and the smaller the color mean value, which indicates that the current particulate matter difference is small and the color vividness value is large.

[0034] A particulate matter image is acquired, and a local magnification is performed to obtain several grids. The number and area of ​​each grid are counted. When a crack is detected in a grid, it is marked as a crack, and the crack area is calculated. The total crack area is obtained by summing all crack areas. The grid area s, the number of cracks nu, and the total crack area sm are then used to calculate the total crack area using a preset formula. The apparent value BGZ is obtained; where b1 and b2 are preset weighting factors, and β is a preset correction coefficient.

[0035] Obtain the apparent values ​​of all particulate matter in the test sample, arrange the apparent values ​​in descending order, and take the median and mode apparent values. It should be noted that the median apparent value represents the apparent value that is in the middle position when arranged from largest to smallest in a set of apparent values; the mode apparent value refers to the value that appears most frequently in a set of apparent values.

[0036] The median apparent value ZSB and the mode apparent value BGZj, where j = 1, 2, 3, ..., n3, i ∈ j and n3 ∈ n2, are expressed using a pre-defined formula. The average crack value LJZ is obtained, where d1 and d2 are preset weighting coefficients. The smaller the difference between the mode apparent value and the median apparent value, the larger the average apparent value, and the larger the average crack value, indicating that the current particulate matter sample has fewer overall cracks and better quality.

[0037] Please see Figure 2 As shown, particulate matter images are acquired, and an image analyzer identifies the edge positions of the particles. A two-dimensional coordinate system is constructed, and points are taken at equal intervals at the edge positions of the particles and marked as edge points. The collected edge points are sorted according to the equal interval length, and edge coordinates are generated from the equal interval length and the edge point height. All edge coordinates are substituted into the two-dimensional coordinate graph to construct a two-dimensional image of the particles. The leftmost and rightmost edge coordinates are connected by a line, and the two-dimensional image of the particles is divided into an upper curve and a lower zigzag curve by this line.

[0038] Divide the region into m equal regions, obtain the height of each point within each region and label it as hr, where r = 1, 2, 3...n4, and n4 is a positive integer. Calculate the average height by averaging the heights of all points within each region. Through preset formula The notch value QKZ is obtained; the larger the notch value, the greater the possibility that the particle has a defect in this area.

[0039] The total gap value is obtained by summing the gap values ​​in all regions of the particulate matter. A preset gap interval Q1 is defined. When the total gap value is greater than the maximum value in the preset gap interval Q1, the particulate matter is marked as severely deficient. When the total gap value is within the preset gap interval Q1, the particulate matter is marked as moderately deficient. When the total gap value is less than the minimum value in the preset gap interval Q1, the particulate matter is marked as poorly deficient.

[0040] The number of particulate matter in the statistical sample was labeled as severely deficient, moderately deficient, and poorly deficient, and then labeled as mu1, mu2, and mu3, respectively. When mu1 ≥ mu2 + mu3, the severely deficient coefficient was matched; when mu3 ≥ mu1 + mu2, the poorly deficient coefficient was matched; and in other cases, the moderately deficient coefficient was matched. The severely deficient coefficient, moderately deficient coefficient, and poorly deficient coefficient were labeled as the residual coefficient Qp, where p = 1, 2, 3, p1 represents the severely deficient coefficient, p2 represents the moderately deficient coefficient, and p3 represents the poorly deficient coefficient.

[0041] The grayscale values ​​of the internal spectrum of the particulate matter are obtained by performing spectral analysis. It should be noted that the grayscale value reflects the density value inside the particulate matter. The larger the grayscale value, the greater the density. When the particulate matter is hollow or contains impurities, the grayscale value inside the particulate matter will be abnormal. A standard grayscale value is preset for each different particulate matter. The grayscale difference is calculated by the difference between the grayscale value and the marked grayscale value.

[0042] The color mean value SJZ, crack mean value LJZ, defect coefficient Qp, and grayscale difference HD are calculated using a preset formula. The standard quality value BZZ is obtained, where f1, f2, f3, and f4 are preset weighting coefficients, and λ is a preset correction coefficient. A preset standard quality value is also obtained. When the standard quality value is greater than the standard quality value, it is recorded as the first sample particulate matter signal and displayed with the text "First particulate matter sample". Otherwise, it is recorded as the second sample particulate matter signal and displayed with the text "Second particulate matter sample". The detection count is incremented by one, and the first sample particulate matter signal and the second sample particulate matter signal are marked as the detection results.

[0043] The server performs a backtracking traversal correction operation, with the following specific steps:

[0044] S1: Establish a first database and a second database. Send the particulate information corresponding to the particulate signal of the first sample to the first database for storage, and send the particulate information corresponding to the particulate signal of the second sample to the second database for storage.

[0045] S2: Select any uncorrected detection result from either the first or second database, select a target detection end, and send particulate information to the selected target detection end for detection and judgment to obtain a correction result; if the received correction result is incorrect, correct and adjust the particulate information and output the detection result again, and record the corresponding detection result as the corrected detection result; if the received correction result is correct, output the first particulate sample prompt and continue to select particulates for detection and judgment to ensure that all detection results in the database are traversed;

[0046] S21: Obtain the detection end corresponding to the detection result, and use the detection count T1, active error correction count T2, and passive error correction count T3 of other detection ends to perform a preset formula. The judgment value TPZ is obtained, where z1, z2 and z are preset weight coefficients, and μ is a preset correction coefficient. The detection end with the largest judgment value is recorded as the target detection end. The target detection end receives particulate matter information, performs detection, and outputs the detection result again.

[0047] S22: Send the re-detection result to the detection end to overwrite the detection result, match the detection result with the re-detection result, if they are inconsistent, generate an incorrect result, if they are consistent, generate a correct result;

[0048] S23: When an uncorrected detection result from the first database is selected, the particulate information corresponding to the selected first particulate sample is stored in the second database; when an uncorrected detection result from the second database is selected, the particulate information corresponding to the selected second particulate sample is stored in the first database, and the corresponding detection end is recorded as one passive error correction. The number of passive error corrections of the detection end is counted. When the number of passive error corrections exceeds the preset number, the detection end stops working and a maintenance notice is issued to the staff; the target detection end is recorded as one active error correction; the selected detection result is corrected, and all particulate information and corresponding detection results in the database are updated and saved; if the detection result is incorrect due to environmental differences and detection instrument errors, the first database and the second database are mutually corrected and adjusted to establish a particulate identification database with correct detection results.

[0049] For example, in this invention, the particulate matter can be grain particles such as rice, soybeans, wheat, and sorghum. If the particulate matter is rice, the parameters and images of the rice are collected.

[0050] The image (rice color and brightness values, image is an internal spectrum of rice, rice image), the parameters of rice and image are processed to obtain color mean, crack mean, defect coefficient and gray level difference, and these are normalized to obtain standard quality value. The standard quality value is compared with the corresponding set threshold to output the rice detection result and display the rice detection result description; the rice detection result description includes rice qualified result or rice unqualified result (particulate matter of the first sample and particulate matter of the second sample).

[0051] The rice detection results and corresponding rice information are sent to the server and saved. The rice information includes rice parameters and rice images. Qualified rice results are entered into the first database, and unqualified rice results are entered into the second database. A backtracking correction operation is performed on all detection results in the first and second databases to obtain the corresponding correction results. When the correction result is incorrect, the rice parameters and images are corrected and adjusted, and a re-detection result is output. The re-detection result is sent to the detection terminal to overwrite the corresponding detection result. The rice parameters, images, and corresponding detection results in all databases are updated and saved.

[0052] In use, this invention processes received particulate matter information into color mean, crack mean, defect coefficient, and grayscale difference, normalizes these values, and analyzes them to obtain a standard quality value. The standard quality value is then compared with a set threshold, and the detection results are output and displayed, enabling multi-faceted and more comprehensive particulate matter detection. By classifying and backtracking the detection results, a suitable detection end is selected for re-detection. If the detection result matches the first one, further random detection results are selected for verification. This ensures that all test results are traversed. When test results are inconsistent, a second test result is output, overwriting the previous result. The test result is recorded as a passive error correction. The number of passive error corrections at the test end is counted. When the number of passive error corrections exceeds a preset number, the test end is taken offline and notified for maintenance. This not only enables backtracking and correction of test results, improving their accuracy, but also takes offline test ends with more than a preset number of errors, thus monitoring the test end and improving the accuracy of test results.

[0053] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A solid particulate matter collection and testing information identification and processing system, comprising a detection terminal and a server, characterized in that, The detection terminal is used to identify solid particulate matter to obtain particulate matter information. The particulate matter information is processed to obtain color mean value, crack mean value, defect coefficient and gray scale difference value. Then, it is analyzed to obtain standard quality value. The standard quality value is compared with the set corresponding threshold, and the detection result is output and displayed. The detection terminal increments the detection count by one and sends the detection result and corresponding particulate matter information to the server. The specific process for generating the incompleteness coefficients is as follows: Acquire particulate matter images, identify the edge positions of the particles, construct a two-dimensional coordinate system, connect the leftmost and rightmost edge coordinates, and use this line to segment the two-dimensional image of the particulate matter into an upper curve and a lower zigzag curve; divide it into m regions of equal length, obtain the height of each point in each region, calculate the average height of the height of each point in each region, and obtain the gap value by processing the height and average height. The total gap value is obtained by summing the gap values ​​in all regions of the particulate matter. The total gap value is then compared with the corresponding set thresholds to classify the particulate matter into severely deficient, moderately deficient, and lightly deficient particles. The corresponding quantities are counted and labeled as mu1, mu2, and mu3, respectively. When mu1 ≥ mu2 + mu3, the severely deficient coefficient is obtained; when mu3 ≥ mu1 + mu2, the lightly deficient coefficient is obtained; otherwise, the moderately deficient coefficient is obtained. The severely deficient, moderately deficient, or lightly deficient coefficients are labeled as the deficiencies Qp, where p = 1, 2, 3, p1 represents the severely deficient coefficient, p2 represents the moderately deficient coefficient, and p3 represents the lightly deficient coefficient. The server establishes a database based on the detection results. The database stores all particulate matter parameters, images, and corresponding detection results uploaded by all detection terminals. Based on the detection results, the corresponding particulate matter parameters and images are classified and stored in different databases. A backtracking and correction operation is performed on the particulate matter information in all databases to obtain the correction results. When the correction result is incorrect, the particulate matter information is corrected and adjusted, and the detection result is output again. The detection result is sent to the detection terminal to overwrite the detection result, and the particulate matter information and corresponding detection results in all databases are updated and saved.

2. The information identification and processing system for solid particulate matter collection and testing according to claim 1, characterized in that, The specific steps for generating the color mean are as follows: The color and brightness values ​​of particulate matter are obtained, particulate matter is identified and the corresponding standard color and brightness values ​​are retrieved from the database, the color and brightness values ​​of particulate matter are normalized and their values ​​are taken, and the values ​​are processed to obtain the color vividness value. The average color vividness value is obtained by retrieving the color vividness values ​​of all particles in the test sample and taking the average value. The average color vividness value is then obtained through data processing.

3. The information identification and processing system for solid particulate matter collection and testing according to claim 2, characterized in that, The specific steps for generating the split mean are as follows: Acquire particulate matter images, divide them into several grids, and count the number and area of ​​each grid. When a crack is detected in a grid, mark it as a crack and calculate the crack area. Sum all the crack areas to obtain the total crack area. Process the grid area, number of cracks, and total crack area to obtain the apparent value. Obtain the apparent values ​​of all particulate matter in the test sample, and select the median apparent value and the mode apparent value; obtain the crack mean value from the median apparent value and the mode apparent value through data processing.

4. The information identification and processing system for solid particulate matter collection and testing according to claim 3, characterized in that, The specific steps for generating grayscale difference are as follows: The grayscale value inside the particulate matter is obtained by performing spectral analysis. A standard grayscale value is preset for each different particulate matter. The difference between the grayscale value and the marked grayscale value is calculated to obtain the grayscale difference value.

5. The information identification and processing system for solid particulate matter collection and testing according to claim 1, characterized in that, The specific steps for the backtracking unit to perform backtracking and correction of the detection results are as follows: S1: Establish a first database and a second database based on the particulate matter detection results, and save the detection results and the corresponding particulate matter information to the first database and the second database respectively; S2: Select any detection result from the first database, select the target detection end and send the corresponding particulate matter information to the target detection end for detection and judgment to obtain the correction result. If the received correction result is incorrect, the particulate matter information will be corrected and adjusted and the detection result will be output again. The detection end is recorded as one passive error correction, the target detection end is recorded as one active error correction, and the corresponding detection result is recorded as the corrected detection result; when the correction result is received as correct, the first particulate matter sample prompt is output and particulate matter is selected for detection and judgment.

6. The information identification and processing system for solid particulate matter collection and testing according to claim 5, characterized in that, The specific steps for selecting the target detection end are as follows: Obtain the detection end corresponding to the detection result, and record the number of detections, active error corrections, and passive error corrections of other detection ends. Normalize these numbers and obtain their values. Process and analyze these values ​​to obtain the judgment value. Record the detection end with the largest judgment value as the target detection end. The target detection end receives particulate matter information, performs particulate matter detection, and outputs the detection result again.

Images