Cat litter quality inspection method and system based on image recognition

Abstract

The present invention relates to the technical field of cat litter quality inspection, and in particular to a cat litter quality inspection method and system based on image recognition, the method comprising: extracting cat litter samples and placing them in a test container for numbering; performing single-point pouring tests with preset amounts of simulated urine at different positions, collecting cat litter sample images in real time, analyzing the actual absorption rate of the cat litter samples, and determining the cat litter absorption capacity; obtaining a number of test images, identifying the agglomeration area area corresponding to agglomeration particles, and determining the cat litter agglomeration capacity according to the comparison result of the agglomeration area area; separating agglomeration particles, shaking the agglomeration particles, analyzing the images before and after shaking, and judging the cat litter disintegration based on the analysis results; generating a quality inspection form based on a number of numbers, cat litter absorption capacity, cat litter agglomeration capacity and cat litter disintegration, analyzing the quality inspection form, and determining the quality inspection report of this batch of cat litters based on the analysis results. The present invention improves the accuracy of the quality inspection results.

Description

Technical Field

[0001] The present invention relates to the technical field of cat litter quality inspection, and in particular to a cat litter quality inspection method and system based on image recognition. Background Art

[0002] The performance evaluation of cat litter mainly includes water absorption, clumping, deodorization, dust content and other aspects. Water absorption is one of the most basic properties of cat litter, which is related to whether the cat litter can quickly absorb cat excrement and keep dry. Clumping refers to whether the cat litter can quickly form solid clumps after absorbing excrement, which is easy to clean.

[0003] The patent document with Chinese patent application number CN111280074A discloses natural plant cat litter and production process. Natural plant cat litter can be divided into original flavor, green tea flavor, lavender flavor, milk flavor, strawberry flavor, peach flavor and other flavors. The ingredients of each flavor are made of natural and different food pigments and flavors according to their corresponding proportions. The production process of the above-mentioned natural plant cat litter is calculated by weight. The raw materials such as dry bean dregs 45%-35%; wheat fiber 25%-15%; edible starch 23%-33%; water 5%-15% are subjected to the following steps: ingredient mixing - uniform stirring - molding and granulation - dehydration and drying - material subdivision - transmission and cooling - quality inspection test - moisture-proof packaging.

[0004] The prior art provides natural plant cat litters with various flavors and detailed production processes thereof. However, in actual production, there is still a problem in ensuring that each batch of cat litter meets the established quality standards. In particular, in the quality inspection of cat litter, simple physical tests are not only inefficient but also prone to errors, resulting in inaccurate quality inspection results. Summary of the invention

[0005] To this end, the present invention provides a cat litter quality inspection method and system based on image recognition, which can solve the problem of inaccurate quality inspection results.

[0006] To achieve the above object, the present invention provides a cat litter quality inspection method based on image recognition, the method comprising: extracting a number of cat litter samples of preset weight from the production batch and evenly laying them in a test container, numbering the cat litter samples respectively, and obtaining a number of numbers;

[0007] Performing a single-point pouring test on a plurality of cat litter samples at different positions in the test container with a preset amount of simulated urine, collecting images of the cat litter samples in real time at preset time intervals, obtaining a plurality of real-time images, analyzing an actual absorption rate of the cat litter samples based on the plurality of real-time images, comparing the actual absorption rate with a preset absorption rate, and determining the absorption capacity of the cat litter according to the comparison result;

[0008] After a preset period of time, a plurality of test images of the cat litter samples are acquired based on an image acquisition device, the plurality of test images are analyzed to identify the area of ​​the clumping region corresponding to the clumping particles, the area of ​​the clumping region is compared with a preset area, and the clumping ability of the cat litter is determined according to the comparison result;

[0009] Separate the agglomerated particles in any of the cat litter samples, shake the agglomerated particles according to a preset speed, analyze the pre-shaking image and the post-shaking image corresponding to the agglomerated particles before and after the shaking, and determine the disintegrability of the cat litter based on the analysis results;

[0010] A quality inspection form is generated based on the number, the cat litter absorption capacity, the cat litter clumping ability and the cat litter disintegration, the quality inspection form is analyzed, and a quality inspection report for this batch of cat litters is determined based on the analysis results.

[0011] Furthermore, the step of performing a single-point pouring test on a plurality of cat litter samples at different positions in the test container with a preset amount of simulated urine comprises:

[0012] Taking the center of the test container as a reference, the test container is divided into left and right sides at preset intervals to obtain a plurality of areas to be tested;

[0013] Putting cat simulated urine into the test simulation device, and performing tests on several of the test areas based on the test simulation device simulating the cat's urination method, until the simulated urine is exhausted.

[0014] Further, the step of analyzing the actual absorption rate of the cat litter sample based on a plurality of real-time images comprises:

[0015] Identify a number of brightness values ​​in the real-time image, mark an area whose brightness value is less than a preset brightness value, and obtain a marked area;

[0016] The area difference of the marked area in the real-time images taken at adjacent preset time intervals is calculated, and the actual absorption rate is calculated based on the area difference and the preset time interval.

[0017] Furthermore, the step of calculating the area difference of the marked area in the real-time images taken at adjacent times includes:

[0018] Identifying a marking boundary of the marking area based on an edge contour algorithm;

[0019] Traversing the area surrounded by the marked boundary based on a preset pixel block to obtain a plurality of sub-region blocks, and counting the number of the plurality of sub-region blocks to obtain a first value;

[0020] Counting the number of sub-region blocks of the marked region obtained at adjacent preset time intervals to obtain a second value;

[0021] The numerical difference between the first numerical value and the second numerical value is calculated, and the numerical difference is multiplied by a preset pixel block to obtain the area difference.

[0022] Furthermore, the step of analyzing the plurality of test images to identify the agglomerated region areas corresponding to the agglomerated particles comprises:

[0023] Extracting texture features corresponding to the test image based on a texture analysis algorithm;

[0024] Identify the actual clustering boundary corresponding to the clustering area based on the texture feature and the edge detection algorithm;

[0025] The area of ​​the clustering region is calculated according to the actual clustering boundary.

[0026] Furthermore, the step of analyzing the pre-shaking image and the post-shaking image corresponding to the agglomerated particles before and after the shaking includes:

[0027] Respectively identifying a number of edge contours of the image before and after the shake based on an edge detection algorithm;

[0028] Converting the pre-shaking image and the post-shaking image into binary images, and distinguishing a gap area and a particle area based on a plurality of edge contours and the binary images;

[0029] Counting a first number and a first total area of ​​gap regions corresponding to the pre-shake image;

[0030] The second number and the second total area of ​​the gap regions corresponding to the shaken image are counted.

[0031] Further, the step of determining the disintegrability of cat litter based on the analysis result comprises:

[0032] comparing the first quantity with the second quantity, and calculating an actual quantity change rate based on the comparison result;

[0033] comparing the first total area with the second total area, and calculating an actual area change rate based on the comparison result;

[0034] Compare the actual quantity change rate with a preset quantity change threshold to obtain a first comparison result;

[0035] Compare the actual area change rate with a preset area change threshold to obtain a second comparison result;

[0036] The disintegrability of the cat litter is determined based on the first comparison result and the second comparison result.

[0037] Furthermore, the step of identifying the actual clustering boundary corresponding to the clustering area based on the texture feature and the edge detection algorithm includes:

[0038] Determining an initial clustering boundary based on the comparison result of the texture features;

[0039] The edge information in the test image is identified based on the edge detection algorithm, and the initial clumping boundary is corrected based on the edge information to obtain the actual clumping boundary.

[0040] Further, the step of determining the quality inspection report of this batch of cat litter based on the analysis result includes:

[0041] Determine a number of serial numbers of cat litter samples that meet the quality inspection standard based on the quality inspection form;

[0042] Count the number of several numbers, compare them with the total number of numbers, and generate a quality inspection report based on the comparison results.

[0043] On the other hand, the present invention also provides a system for a cat litter quality inspection method based on image recognition, the system comprising:

[0044] A numbering module is used to extract a number of cat litter samples of preset weight from the production batch and evenly spread them into a test container, and to number the cat litter samples respectively to obtain a number of numbers;

[0045] an absorption detection module, for performing a single-point pouring test on a plurality of cat litter samples at different positions in the test container with a preset amount of simulated urine, collecting images of the cat litter samples in real time at preset time intervals, obtaining a plurality of real-time images, analyzing an actual absorption rate of the cat litter samples based on the plurality of real-time images, comparing the actual absorption rate with a preset absorption rate, and determining the absorption capacity of the cat litter according to the comparison result;

[0046] a clumping determination module, used to obtain a plurality of test images of the cat litter samples based on an image acquisition device after a preset period of time, analyze the test images to identify the clumping area corresponding to the clumping particles, compare the clumping area with a preset area, and determine the clumping ability of the cat litter according to the comparison result;

[0047] A disintegration determination module is used to separate the agglomerated particles in any cat litter sample, shake the agglomerated particles according to a preset speed, analyze the pre-shaking image and the post-shaking image corresponding to the agglomerated particles before and after the shaking, and determine the disintegration of the cat litter based on the analysis results;

[0048] The report generation module is used to generate a quality inspection form based on the number, the cat litter absorption capacity, the cat litter clumping ability and the cat litter disintegration, analyze the quality inspection form, and determine the quality inspection report of this batch of cat litters based on the analysis results.

[0049] Compared with the prior art, the beneficial effects of the present invention are that, by extracting cat litter samples of preset weight and spreading them evenly, it is ensured that each sample has the same initial condition at the beginning of the test, thereby improving the comparability and accuracy of the test results; by numbering the cat litter samples, each sample can be accurately identified and tracked in subsequent tests and analyses, thereby enhancing the traceability of the quality inspection process; by collecting images of cat litter samples in real time at preset time intervals, the dynamic process of cat litter absorbing urine can be captured, thereby more accurately analyzing the actual absorption rate; based on the real-time image analysis of the actual absorption rate of the cat litter sample and comparing it with the preset absorption rate, the absorption capacity of the cat litter can be accurately evaluated, and single-point analysis at different positions in the test container can be performed. The pouring liquid simulated urine test evaluates the absorption performance of the cat litter samples at different positions, which increases the comprehensiveness of the test. The test images are obtained through the image acquisition equipment, which intuitively shows the morphology of the cat litter after clumping, which is convenient for subsequent analysis and evaluation. By identifying the clumping area corresponding to the clumping particles and comparing it with the preset area, the quantitative evaluation of the clumping ability of the cat litter is realized, which improves the accuracy and objectivity of the evaluation. By shaking the clumping particles and analyzing the images before and after shaking, the disintegration of the cat litter when subjected to external forces can be accurately evaluated, providing an accurate data basis for the subsequent cat litter quality inspection and evaluation. The quality of this batch of cat litter is comprehensively and comprehensively evaluated based on the absorption capacity, clumping ability and disintegration, which helps to make the cat litter quality inspection and evaluation results more accurate.

[0050] In particular, image processing technology is used to accurately identify and mark areas with brightness values ​​below a preset threshold, so that subtle changes in the absorption process can be captured and analyzed. Images are collected and analyzed in real time at preset time intervals, which can reflect changes in the absorption rate of the cat litter sample in real time, making the test results more accurate. The actual absorption rate is calculated based on the area difference and the preset time interval, making the calculated absorption rate data more accurate and objective, which helps to accurately evaluate the absorption performance of the cat litter.

[0051] In particular, the boundary of the marked area is accurately identified through the edge contour algorithm to ensure the accuracy of subsequent area calculation. The marked area is subdivided into several sub-area blocks, and the area difference is calculated by counting the number of sub-area blocks, which improves the accuracy of the calculation. The preset pixel block is used as the reference unit for area calculation, which reduces the error caused by changes in image resolution or size. The area difference is calculated through the difference in quantity, which avoids the error caused by direct measurement of the area, and helps to verify the stability and reliability of the test results.

[0052] In particular, the texture features in the image are captured through the texture analysis algorithm, and by accurately extracting these features, the clustering area can be more effectively identified, thereby improving the accuracy of recognition. The edge detection algorithm is used to identify the boundary information in the image. Combined with the texture features, the edge detection algorithm can more accurately locate the actual clustering boundaries of the clustered particles, thereby improving the measurement accuracy and providing a reliable basis for subsequent area calculations. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A schematic diagram of a process for a cat litter quality inspection method based on image recognition provided by an embodiment of the present invention;

[0054] Figure 2 A schematic structural diagram of a testing device for a cat litter quality inspection method based on image recognition provided by an embodiment of the present invention;

[0055] Figure 3 A schematic diagram of a process for identifying the area of ​​a clumping region in a cat litter quality inspection method based on image recognition provided by an embodiment of the present invention;

[0056] Figure 4 A structural block diagram of a cat litter quality inspection system based on image recognition provided by an embodiment of the present invention;

[0057] Figure 5 A schematic diagram of cat litter clumping in a cat litter quality inspection method based on image recognition provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0059] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0060] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it should not be understood as a limitation on the present invention.

[0061] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0062] See also Figure 1 As shown, an embodiment of the present invention provides a cat litter quality inspection method based on image recognition, the method comprising:

[0063] Step S100, extracting a number of cat litter samples of preset weight from the production batch and evenly laying them into a test container, numbering the cat litter samples respectively, and obtaining a number of numbers;

[0064] Step S200, performing a single-point pouring test on a plurality of cat litter samples at different positions in the test container with a preset amount of simulated urine, collecting images of the cat litter samples in real time at preset time intervals, obtaining a plurality of real-time images, analyzing an actual absorption rate of the cat litter samples based on the plurality of real-time images, comparing the actual absorption rate with a preset absorption rate, and determining the absorption capacity of the cat litter according to the comparison result;

[0065] Step S300, acquiring a plurality of test images of the cat litter samples based on an image acquisition device after a preset period of time, analyzing the plurality of test images to identify the agglomeration area corresponding to the agglomerated particles, comparing the agglomeration area with a preset area, and determining the agglomeration ability of the cat litter according to the comparison result;

[0066] Step S400, separating the agglomerated particles in any of the cat litter samples, shaking the agglomerated particles according to a preset speed, analyzing the images before and after shaking corresponding to the agglomerated particles, and determining the disintegrability of the cat litter based on the analysis results;

[0067] Step S500, generating a quality inspection form based on the serial numbers, the cat litter absorption capacity, the cat litter clumping capacity and the cat litter disintegration, analyzing the quality inspection form, and determining a quality inspection report for this batch of cat litters based on the analysis results.

[0068] Specifically, the embodiment of the present invention extracts cat litter samples of preset weight and spreads them evenly, thereby ensuring that each sample has the same initial condition at the beginning of the test, thereby improving the comparability and accuracy of the test results. By numbering the cat litter samples, each sample can be accurately identified and tracked in subsequent tests and analyses, thereby enhancing the traceability of the quality inspection process. By collecting images of cat litter samples in real time at preset time intervals, the dynamic process of cat litter absorbing urine can be captured, thereby more accurately analyzing the actual absorption rate. Based on the real-time image analysis of the actual absorption rate of the cat litter sample and comparing it with the preset absorption rate, the absorption capacity of the cat litter can be accurately evaluated, and single-point pouring simulation can be performed at different positions in the test container. The urine test evaluates the absorption performance of cat litter samples at different positions, which increases the comprehensiveness of the test. The test images are obtained through image acquisition equipment, which intuitively displays the morphology of the cat litter after clumping, facilitating subsequent analysis and evaluation. By identifying the clumping area corresponding to the clumping particles and comparing it with the preset area, a quantitative evaluation of the clumping ability of the cat litter is achieved, which improves the accuracy and objectivity of the evaluation. By shaking the clumping particles and analyzing the images before and after shaking, the disintegration of the cat litter when subjected to external forces can be accurately evaluated, providing an accurate data basis for the subsequent cat litter quality inspection and evaluation. The quality of this batch of cat litter is comprehensively and comprehensively evaluated based on the absorption capacity, clumping ability and disintegration, which helps to make the cat litter quality inspection and evaluation results more accurate.

[0069] Specifically, the preset weight in the embodiment of the present invention is the weight when the cat litter sample is spread flat in the test container and occupies 1 / 3 of the test container;

[0070] The preset time interval is 2s;

[0071] The preset time period is based on the average of cat litter clumping time during historical testing;

[0072] Specifically, the step of performing a single-point pouring test on a plurality of cat litter samples at different positions in the test container with a preset amount of simulated urine comprises:

[0073] Taking the center of the test container as a reference, the test container is divided into left and right sides at preset intervals to obtain a plurality of areas to be tested;

[0074] Putting cat simulated urine into the test simulation device, and performing tests on several of the test areas based on the test simulation device simulating the cat's urination method, until the simulated urine is exhausted.

[0075] Specifically, the preset interval in the embodiment of the present invention is 1 / 10 of the length of the test container.

[0076] See also Figure 2As shown, it is a structural schematic diagram of the testing equipment of the cat litter quality inspection method based on image recognition according to an embodiment of the present invention, wherein, a support frame 1, a telescopic rod 2, a driving motor 3, a push rod 4, an extrusion plate 5, a capsule storage bin 6, a derivation tube 7 and an outlet 8, the support frame 1 is fixedly connected to the telescopic rod 2, the other end of the telescopic rod 2 is fixedly connected to the extrusion plate 5, the driving motor 3 is fixedly connected to the push rod 4, the push rod 4 is fixedly connected to the extrusion plate 5, the extrusion plate 5 is fitted with the capsule storage bin 6, the derivation tube 7 is fixedly connected to the capsule storage bin 6 and forms a certain angle with the horizontal axis of the capsule storage bin 6, and the outlet 8 is fixedly connected to the derivation tube 7.

[0077] Specifically, the certain angle in the embodiment of the present invention is an angle arbitrarily selected between 20° and 40°.

[0078] Specifically, the steps of simulating the urination mode of a cat based on the test simulation device in the embodiment of the present invention include:

[0079] The driving motor 3 controls the push rod 4 to move at the first preset speed, and the push rod 4 controls the squeezing plate 5 to squeeze the capsule storage bin 6, so that the simulated cat urine in the capsule storage bin 6 flows out to the area to be tested through the outlet pipe 7 and the outlet 8;

[0080] When 1 / 5 of the simulated cat urine remains in the capsule storage compartment 6, the first preset speed is adjusted to the second preset speed, and the push rod 4 is continuously controlled to move until the simulated cat urine in the capsule storage compartment 6 is completely drained and stops.

[0081] Specifically, the step of analyzing the actual absorption rate of the cat litter sample based on a plurality of real-time images comprises:

[0082] Identify a number of brightness values ​​in the real-time image, mark an area whose brightness value is less than a preset brightness value, and obtain a marked area;

[0083] The area difference of the marked area in the real-time images taken at adjacent preset time intervals is calculated, and the actual absorption rate is calculated based on the area difference and the preset time interval.

[0084] Specifically, the embodiment of the present invention uses image processing technology to accurately identify and mark areas with brightness values ​​below a preset threshold, thereby capturing and analyzing subtle changes in the absorption process, and collecting and analyzing images in real time at preset time intervals. It can reflect changes in the absorption rate of cat litter samples in real time, making the test results more accurate, and calculating the actual absorption rate based on the area difference and the preset time interval, so that the calculated absorption rate data is more accurate and objective, which helps to accurately evaluate the absorption performance of cat litter.

[0085] Specifically, the step of identifying a plurality of brightness values ​​in the real-time image in the embodiment of the present invention includes:

[0086] Converting the real-time images into grayscale images;

[0087] A plurality of pixel values ​​of the grayscale image are traversed and identified, and a plurality of brightnesses are obtained based on the plurality of pixel values.

[0088] Specifically, the actual absorption rate in the embodiment of the present invention is the area difference divided by the time interval.

[0089] Specifically, the step of determining the cat litter absorption capacity according to the comparison result in the embodiment of the present invention includes:

[0090] When the actual absorption rate is greater than or equal to the preset absorption rate, the cat litter has a strong absorption capacity;

[0091] When the actual absorption rate is less than the preset absorption rate, the cat litter has a weak absorption capacity.

[0092] Specifically, the preset absorption rate described in the embodiment of the present invention is the average absorption rate of clumping cat litter samples in several historical qualified cat litter batches.

[0093] Specifically, an embodiment of the present invention traverses and identifies several pixel values ​​of the grayscale image by using the cv2.imread function of OpenCV to read the image, and specifies cv2.IMREAD_GRAYSCALE to ensure that the image is converted into a grayscale image, and uses a nested for loop to traverse each pixel in the image. In each iteration, image[y,x] is used to access the brightness value of the pixel and store it in the list brightness_values.

[0094] Specifically, the step of calculating the area difference of the marked area in the real-time images taken at adjacent times includes:

[0095] Identifying a marking boundary of the marking area based on an edge contour algorithm;

[0096] Traversing the area surrounded by the marked boundary based on a preset pixel block to obtain a plurality of sub-region blocks, and counting the number of the plurality of sub-region blocks to obtain a first value;

[0097] Counting the number of sub-region blocks of the marked region obtained at adjacent preset time intervals to obtain a second value;

[0098] The numerical difference between the first numerical value and the second numerical value is calculated, and the numerical difference is multiplied by a preset pixel block to obtain the area difference.

[0099] Specifically, the embodiment of the present invention accurately identifies the boundary of the marked area through an edge contour algorithm to ensure the accuracy of subsequent area calculation, subdivides the marked area into a number of sub-area blocks, and calculates the area difference by counting the number of sub-area blocks, thereby improving the accuracy of the calculation. A preset pixel block is used as the reference unit for area calculation, which reduces the error caused by changes in image resolution or size. The area difference is calculated through quantitative differences, which avoids the error caused by direct area measurement, and helps to verify the stability and reliability of the test results.

[0100] Specifically, the preset pixel block in the embodiment of the present invention is 16x16 pixels.

[0101] Specifically, at time point T1, the embodiment of the present invention counts that the number of sub-region blocks completely located in the marked area is 100. At the next preset time interval, i.e., time point T2, the above process is repeated and the number of sub-region blocks is 90. The calculated area difference is: (100-90)×(16×16)=960 pixels. The number of pixels is converted into actual area units (such as square centimeters) according to the resolution of the image (i.e., the number of pixels per inch or per centimeter).

[0102] See also Figure 3 As shown, the step of analyzing the plurality of test images to identify the agglomerated area corresponding to the agglomerated particles comprises:

[0103] Step S310, extracting texture features corresponding to the test image based on a texture analysis algorithm;

[0104] Step S320, identifying the actual clustering boundary corresponding to the clustering area based on the texture feature and the edge detection algorithm;

[0105] Step S330: Calculate the area of ​​the clustering region according to the actual clustering boundary.

[0106] Specifically, the embodiments of the present invention capture texture features in the image through a texture analysis algorithm, and more effectively identify the clumping area by accurately extracting these features, thereby improving the accuracy of recognition. The edge detection algorithm is used to identify the boundary information in the image. Combined with the texture features, the edge detection algorithm can more accurately locate the actual clumping boundaries of the clumping particles, thereby improving the measurement accuracy and providing a reliable basis for subsequent area calculations.

[0107] Specifically, the texture analysis algorithm described in the embodiment of the present invention includes gray-level co-occurrence matrix, local binary pattern (LBP), Gabor filter, etc.;

[0108] The edge detection algorithm may be a Canny edge detection algorithm.

[0109] Specifically, an embodiment of the present invention extracts texture features corresponding to the test image based on a texture analysis algorithm by applying the GLCM algorithm to the test image, calculates the grayscale dependencies in different directions and distances, generates a grayscale co-occurrence matrix, and extracts texture features such as energy, contrast, and entropy from the grayscale co-occurrence matrix.

[0110] Specifically, the embodiment of the present invention can calculate the area of ​​the clumping region according to the actual clumping boundary by utilizing the area calculation function in the image processing software or algorithm library, counting the number of pixels marked as the clumping region, and converting the number of pixels into actual area units (such as square centimeters) according to the resolution of the image (i.e., the number of pixels per inch or per centimeter).

[0111] Specifically, the embodiment of the present invention determines the clumping ability of cat litter according to the comparison result, including:

[0112] When the clumping area is greater than or equal to the preset area, the cat litter has a strong clumping ability;

[0113] When the clumping area is smaller than the preset area, the clumping ability of the cat litter is weak.

[0114] Specifically, the preset area described in the embodiment of the present invention is the average area of ​​clumping cat litter samples in several historical qualified cat litter batches.

[0115] Specifically, the step of analyzing the pre-shaking image and the post-shaking image corresponding to the agglomerated particles before and after the shaking comprises:

[0116] Respectively identifying a number of edge contours of the image before and after the shake based on an edge detection algorithm;

[0117] Converting the pre-shaking image and the post-shaking image into binary images, and distinguishing a gap area and a particle area based on a plurality of edge contours and the binary images;

[0118] Counting a first number and a first total area of ​​gap regions corresponding to the pre-shake image;

[0119] The second number and the second total area of ​​the gap regions corresponding to the shaken image are counted.

[0120] Specifically, after the embodiment of the present invention converts the pre-shake image and the post-shake image into binary images (ie, black and white images), the particle area is white, and the gap area is black.

[0121] Specifically, the step of judging the disintegrability of cat litter based on the analysis result comprises:

[0122] comparing the first quantity with the second quantity, and calculating an actual quantity change rate based on the comparison result;

[0123] comparing the first total area with the second total area, and calculating an actual area change rate based on the comparison result;

[0124] Compare the actual quantity change rate with a preset quantity change threshold to obtain a first comparison result;

[0125] Compare the actual area change rate with a preset area change threshold to obtain a second comparison result;

[0126] The disintegrability of the cat litter is determined based on the first comparison result and the second comparison result.

[0127] Specifically, the number change rate and area change rate described in the embodiment of the present invention can be calculated by the following formula:

[0128] Quantity change rate = (second quantity - first quantity) / first quantity × 100%

[0129] Area change rate = (second total area - first total area) / first total area × 100%.

[0130] Specifically, the preset quantity change threshold in the embodiment of the present invention is 5%;

[0131] The preset area change threshold is 7%.

[0132] Specifically, in an embodiment of the present invention, judging the disintegrability of the cat litter based on the first comparison result and the second comparison result includes:

[0133] When the actual number change rate is greater than or equal to the preset number change threshold and the actual area change rate is greater than or equal to the preset area change threshold, the cat litter has strong disintegration properties;

[0134] When the actual number change rate is greater than or equal to the preset number change threshold or the actual area change rate is greater than or equal to the preset area change threshold, the cat litter is medium in disintegration;

[0135] When the actual quantity change rate is less than the preset quantity change threshold or the actual area change rate is less than the preset area change threshold, the cat litter has weak disintegrability.

[0136] Specifically, the step of identifying the actual clustering boundary corresponding to the clustering area based on the texture feature and the edge detection algorithm includes:

[0137] Determining an initial clustering boundary based on the comparison result of the texture features;

[0138] The edge information in the test image is identified based on the edge detection algorithm, and the initial clumping boundary is corrected based on the edge information to obtain the actual clumping boundary.

[0139] Specifically, the step of determining the initial clustering boundary based on the comparison result of the texture features in the embodiment of the present invention includes:

[0140] Comparing the texture feature with the preset feature to obtain a similarity comparison result;

[0141] The initial clustering boundary is determined based on the similarity comparison result.

[0142] Specifically, the preset features can be obtained by analyzing and counting the texture features of known clumping cat litter samples;

[0143] Similarity comparison can be calculated by using Euclidean distance, cosine similarity, etc.

[0144] Specifically, the step of determining the quality inspection report of this batch of cat litter based on the analysis results includes:

[0145] Determine a number of serial numbers of cat litter samples that meet the quality inspection standard based on the quality inspection form;

[0146] Count the number of several numbers, compare them with the total number of numbers, and generate a quality inspection report based on the comparison results.

[0147] Specifically, the conditions that meet the quality inspection form described in the embodiment of the present invention are that the cat litter has strong absorption capacity, strong clumping ability, and weak disintegration of the cat litter.

[0148] Specifically, the step of generating a quality inspection report based on the comparison result in the embodiment of the present invention includes:

[0149] When the ratio of the number of cat litter samples that meet the quality inspection standard to the total number of cat litter samples is greater than or equal to the preset qualified ratio, a quality inspection report indicating that the batch of cat litter is qualified is generated;

[0150] When the proportion of the number of cat litter sample numbers that meet the quality inspection standards to the total number of numbers is less than the preset qualified proportion, a quality inspection report indicating that this batch of cat litter is unqualified is generated, and the specific reasons for the unqualified are indicated, such as weak clumping ability, strong disintegration, etc.

[0151] Specifically, the preset qualified ratio described in the embodiment of the present invention is 90%.

[0152] Specifically, the quality inspection table described in the embodiment of the present invention is shown in the following table:

[0153] Table 1 Quality inspection form

[0154]

[0155] See also Figure 4 As shown, an embodiment of the present invention also provides a system for a cat litter quality inspection method based on image recognition, the system comprising:

[0156] A numbering module 10 is used to extract a number of cat litter samples of preset weight from the production batch and evenly spread them into a test container, and number the cat litter samples respectively to obtain a number of numbers;

[0157] The absorption detection module 20 is used to perform a single-point pouring test on a plurality of cat litter samples at different positions in the test container with a preset amount of simulated urine, collect images of the cat litter samples in real time at preset time intervals, obtain a plurality of real-time images, analyze the actual absorption rate of the cat litter samples based on the plurality of real-time images, compare the actual absorption rate with the preset absorption rate, and determine the absorption capacity of the cat litter according to the comparison result;

[0158] a clumping determination module 30, for acquiring a plurality of test images of the cat litter samples based on an image acquisition device after a preset period of time, analyzing the plurality of test images to identify the clumping area corresponding to the clumping particles, comparing the clumping area with a preset area, and determining the clumping ability of the cat litter according to the comparison result;

[0159] A disintegration determination module 40 is used to separate the agglomerated particles in any of the cat litter samples, shake the agglomerated particles according to a preset speed, analyze the pre-shaking image and the post-shaking image corresponding to the agglomerated particles before and after the shaking, and determine the disintegration of the cat litter based on the analysis results;

[0160] The report generation module 50 is used to generate a quality inspection form based on the serial numbers, the cat litter absorption capacity, the cat litter clumping capacity and the cat litter disintegration, analyze the quality inspection form, and determine the quality inspection report of this batch of cat litters based on the analysis results.

[0161] Specifically, a system for a cat litter quality inspection method based on image recognition provided in an embodiment of the present invention can execute the above-mentioned cat litter quality inspection method based on image recognition to achieve the same technical effect, which will not be repeated here.

[0162] See also Figure 5 As shown, a schematic diagram of cat litter clumping of a cat litter quality inspection method based on image recognition provided by an embodiment of the present invention is provided. In the figure, granular cat litter and clumping cat litter are included. It can be seen from the figure that the grayscale value of the area corresponding to the clumping cat litter is greater than the grayscale value of the granular cat litter area, so the clumping area is determined.

[0163] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0164] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A cat litter quality inspection method based on image recognition, characterized in that, include: Taking a number of cat litter samples of preset weight from the production batch and evenly laying them in a test container, numbering the cat litter samples respectively, and obtaining a number of numbers; Performing a single-point pouring test on a plurality of cat litter samples at different positions in the test container with a preset amount of simulated urine, collecting images of the cat litter samples in real time at preset time intervals, obtaining a plurality of real-time images, analyzing an actual absorption rate of the cat litter samples based on the plurality of real-time images, comparing the actual absorption rate with a preset absorption rate, and determining the absorption capacity of the cat litter according to the comparison result; After a preset period of time, a plurality of test images of the cat litter samples are acquired based on an image acquisition device, the plurality of test images are analyzed to identify the area of ​​the clumping region corresponding to the clumping particles, the area of ​​the clumping region is compared with a preset area, and the clumping ability of the cat litter is determined according to the comparison result; Separate the agglomerated particles in any of the cat litter samples, shake the agglomerated particles according to a preset speed, analyze the pre-shaking image and the post-shaking image corresponding to the agglomerated particles before and after the shaking, and determine the disintegrability of the cat litter based on the analysis results; A quality inspection form is generated based on the number, the cat litter absorption capacity, the cat litter clumping ability and the cat litter disintegration, the quality inspection form is analyzed, and a quality inspection report for this batch of cat litters is determined based on the analysis results.

2. The cat litter quality inspection method based on image recognition according to claim 1, characterized in that, The step of performing a single-point pouring test on a plurality of cat litter samples at different positions in the test container with a preset amount of simulated urine comprises: Taking the center of the test container as a reference, the test container is divided into left and right sides at preset intervals to obtain a plurality of areas to be tested; Putting cat simulated urine into the test simulation device, and performing tests on several of the test areas based on the test simulation device simulating the cat's urination method, until the simulated urine is exhausted.

3. The cat litter quality inspection method based on image recognition according to claim 2, characterized in that, The step of analyzing the actual absorption rate of the cat litter sample based on a plurality of real-time images comprises: Identify a number of brightness values ​​in the real-time image, mark an area whose brightness value is less than a preset brightness value, and obtain a marked area; The area difference of the marked area in the real-time images taken at adjacent preset time intervals is calculated, and the actual absorption rate is calculated based on the area difference and the preset time interval.

4. The cat litter quality inspection method based on image recognition according to claim 3, characterized in that, The step of calculating the area difference of the marked area in the real-time images taken at adjacent preset time intervals comprises: Identifying a marking boundary of the marking area based on an edge contour algorithm; Traversing the area surrounded by the marked boundary based on a preset pixel block to obtain a plurality of sub-region blocks, and counting the number of the plurality of sub-region blocks to obtain a first value; Counting the number of sub-region blocks of the marked region obtained at adjacent preset time intervals to obtain a second value; The numerical difference between the first numerical value and the second numerical value is calculated, and the numerical difference is multiplied by a preset pixel block to obtain the area difference.

5. The cat litter quality inspection method based on image recognition according to claim 4, characterized in that, The step of analyzing the plurality of test images to identify the agglomerated regions corresponding to the agglomerated particles comprises: Extracting texture features corresponding to the test image based on a texture analysis algorithm; Identify the actual clustering boundary corresponding to the clustering area based on the texture feature and the edge detection algorithm; The area of ​​the clustering region is calculated according to the actual clustering boundary.

6. The cat litter quality inspection method based on image recognition according to claim 5, characterized in that, The step of analyzing the pre-shaking image and the post-shaking image corresponding to the agglomerated particles before and after the shaking comprises: Respectively identifying a number of edge contours of the image before and after the shake based on an edge detection algorithm; Converting the pre-shaking image and the post-shaking image into binary images, and distinguishing a gap area and a particle area based on a plurality of edge contours and the binary images; Counting a first number and a first total area of ​​gap regions corresponding to the pre-shake image; The second number and the second total area of ​​the gap regions corresponding to the shaken image are counted.

7. The cat litter quality inspection method based on image recognition according to claim 6, characterized in that, The step of judging the disintegrability of cat litter based on the analysis result comprises: comparing the first quantity with the second quantity, and calculating an actual quantity change rate based on the comparison result; comparing the first total area with the second total area, and calculating an actual area change rate based on the comparison result; Compare the actual quantity change rate with a preset quantity change threshold to obtain a first comparison result; Compare the actual area change rate with a preset area change threshold to obtain a second comparison result; The disintegrability of the cat litter is determined based on the first comparison result and the second comparison result.

8. The cat litter quality inspection method based on image recognition according to claim 7, characterized in that, The step of identifying the actual clustering boundary corresponding to the clustering area based on the texture feature and the edge detection algorithm comprises: Determining an initial clustering boundary based on the comparison result of the texture features; The edge information in the test image is identified based on the edge detection algorithm, and the initial clumping boundary is corrected based on the edge information to obtain the actual clumping boundary.

9. The cat litter quality inspection method based on image recognition according to claim 8, characterized in that, The step of determining the quality inspection report of this batch of cat litter based on the analysis result comprises: Determine a number of serial numbers of cat litter samples that meet the quality inspection standard based on the quality inspection form; Count the number of several numbers, compare them with the total number of numbers, and generate a quality inspection report based on the comparison results.

10. A system for the cat litter quality inspection method based on image recognition according to any one of claims 1 to 9, characterized in that: include: A numbering module is used to extract a number of cat litter samples of preset weight from the production batch and evenly spread them into a test container, and to number the cat litter samples respectively to obtain a number of numbers; an absorption detection module, for performing a single-point pouring test on a plurality of cat litter samples at different positions in the test container with a preset amount of simulated urine, collecting images of the cat litter samples in real time at preset time intervals, obtaining a plurality of real-time images, analyzing an actual absorption rate of the cat litter samples based on the plurality of real-time images, comparing the actual absorption rate with a preset absorption rate, and determining the absorption capacity of the cat litter according to the comparison result; a clumping determination module, used to obtain a plurality of test images of the cat litter samples based on an image acquisition device after a preset period of time, analyze the test images to identify the clumping area corresponding to the clumping particles, compare the clumping area with a preset area, and determine the clumping ability of the cat litter according to the comparison result; A disintegration determination module is used to separate the agglomerated particles in any cat litter sample, shake the agglomerated particles according to a preset speed, analyze the pre-shaking image and the post-shaking image corresponding to the agglomerated particles before and after the shaking, and determine the disintegration of the cat litter based on the analysis results; The report generation module is used to generate a quality inspection form based on the number, the cat litter absorption capacity, the cat litter clumping ability and the cat litter disintegration, analyze the quality inspection form, and determine the quality inspection report of this batch of cat litters based on the analysis results.

Images