Backfat thickness detection method and device

Abstract

The invention discloses a backfat thickness detection method and device. The method comprises the following steps: acquiring a back image of a pig; inputting the back image into a preset yolk detection model to extract a backfat region image in the back image; performing binarization processing on the backfat area image to obtain a binarized image; the number of image pixels in the binarized image corresponding to the backfat area is calculated, the thickness of the backfat is obtained through a preset mapping function, and the mapping function is established between the thickness value of the backfat and the number of the image pixels in advance. The back fat thickness of the pig can be accurately detected, and the method has wide applicability.

Description

Technical Field

[0001] The present application relates to the technical field of thickness detection, and in particular to a backfat thickness detection method and device. Background Art

[0002] Back fat thickness usually refers to the thickness of subcutaneous fat at a specific position on the pig's back (such as between the fifth and sixth thoracic vertebrae or between the third and fourth ribs from the bottom, and a certain distance from the midline of the back). In the livestock and slaughtering industry, back fat thickness is an important parameter indicator for assessing pork grade. It is a parameter that companies must measure during the production process and can directly reflect the grade of pig carcasses.

[0003] At present, the main method used at home and abroad is to collect back fat images of living pigs through veterinary B-ultrasound instruments. The back fat thickness of living pigs is measured by drawing lines on the back fat images manually. This method is heavily dependent on the experience and skills of the operator and is highly subjective. Therefore, there is a large amount of human error, which in turn affects the accuracy of the measurement. At the same time, manual measurement is time-consuming and inefficient, which makes it difficult to meet large-scale measurement needs and is not widely applicable.

[0004] Application Contents

[0005] The present application provides a backfat thickness detection method and device to accurately detect the backfat thickness of pigs, which has wide applicability.

[0006] In a first aspect, the present application provides a backfat thickness detection method, comprising:

[0007] Obtain an image of the pig's back;

[0008] Inputting the back image into a preset yolo detection model to extract a back fat area image in the back image;

[0009] Binarizing the backfat region image to obtain a binary image;

[0010] The number of image pixels in the binary image corresponding to the backfat area is calculated, and the thickness of the backfat is obtained through a preset mapping function, wherein the mapping function is pre-established between the backfat thickness value and the number of image pixels.

[0011] The present application can accurately extract the back fat area image in the back image through the yolo detection model, so as to facilitate the subsequent accurate calculation of the back fat thickness according to the number of pixels in the back fat area; by binarizing the back fat area image, the center position of the pig's back fat fascia can be better determined, thereby accurately locating the pig's back fat pixels; by calculating the number of image pixels in the binarized image corresponding to the back fat area, and calculating the back fat thickness through a mapping function pre-established between the back fat thickness value and the number of image pixels, the back fat thickness in the back fat area can be accurately detected, and the application has wide applicability.

[0012] Furthermore, the back image is an ultrasound image measured using a veterinary ultrasound instrument.

[0013] In this way, the B-ultrasound image of the pig's back is measured by a veterinary B-ultrasound instrument, which facilitates the subsequent accurate extraction of the back fat area.

[0014] Further, the back image is input into a preset yolo detection model to extract the back fat area image in the back image, specifically:

[0015] Input the back image into a preset yolo detection model to obtain the eye muscle area and back fat area in the back image;

[0016] The back fat region image in the back image is obtained by cropping the back fat region in the back image.

[0017] In this way, the back fat area image in the back image can be accurately extracted through the yolo detection model, which facilitates the subsequent accurate calculation of the back fat thickness according to the number of pixels in the back fat area.

[0018] Furthermore, the training process of the YOLO detection model is as follows:

[0019] Obtain several actual back B-ultrasound images of pigs;

[0020] Using labelme to label actual back fat areas in a plurality of actual back B-ultrasound images to obtain labeling results;

[0021] The annotation results are input into the YOLO detection model to train the YOLO detection model.

[0022] In this way, by training the yolo detection model, the yolo detection model can be used to accurately extract the back fat area image in the back image.

[0023] Furthermore, the number of image pixels in the binary image corresponding to the backfat area is calculated, and the thickness of the backfat is obtained by a preset mapping function, specifically:

[0024] By summing the pixel values ​​of the binary image in the horizontal direction and assigning a pixel value less than a preset value to zero, a peak distribution diagram corresponding to the binary image is obtained;

[0025] Calculate the number of image pixels between the last pixel point of the first peak and the first pixel point of the last peak in the peak distribution graph;

[0026] Based on the number of image pixels, the thickness of the back fat is obtained through a preset mapping function.

[0027] In this way, by calculating the number of image pixels in the binary image corresponding to the backfat area, and calculating the backfat thickness through a mapping function pre-established between the backfat thickness value and the number of image pixels, the backfat thickness in the backfat area can be accurately detected, which has wide applicability.

[0028] In a second aspect, the present application provides a backfat thickness detection device, comprising: an acquisition module, an extraction module, a processing module and a calculation module;

[0029] The acquisition module is used to acquire the back image of the pig;

[0030] The extraction module is used to input the back image into a preset yolo detection model to extract a back fat area image in the back image;

[0031] The processing module is used to perform binarization processing on the back fat area image to obtain a binary image;

[0032] The calculation module is used to calculate the number of image pixels in the binary image corresponding to the backfat area, and obtain the thickness of the backfat through a preset mapping function, wherein the mapping function is pre-established between the backfat thickness value and the number of image pixels.

[0033] The present application can accurately extract the back fat area image in the back image through the yolo detection model, so as to facilitate the subsequent accurate calculation of the back fat thickness according to the number of pixels in the back fat area; by binarizing the back fat area image, the center position of the pig's back fat fascia can be better determined, thereby accurately locating the pig's back fat pixels; by calculating the number of image pixels in the binarized image corresponding to the back fat area, and calculating the back fat thickness through a mapping function pre-established between the back fat thickness value and the number of image pixels, the back fat thickness in the back fat area can be accurately detected, and the application has wide applicability.

[0034] Furthermore, the back image is an ultrasound image measured using a veterinary ultrasound instrument.

[0035] Further, the extraction module includes: an input unit and a clipping unit;

[0036] The input unit is used to input the back image into a preset YOLO detection model to obtain the eye muscle area and back fat area in the back image;

[0037] The cropping unit is used to obtain the back fat area image in the back image by cropping the back fat area in the back image.

[0038] Furthermore, the training process of the YOLO detection model is as follows:

[0039] Obtain several actual back B-ultrasound images of pigs;

[0040] Using labelme to label actual back fat areas in a plurality of actual back B-ultrasound images to obtain labeling results;

[0041] The annotation results are input into the YOLO detection model to train the YOLO detection model.

[0042] Further, the calculation module includes: a first calculation unit, a second calculation unit and a third calculation unit;

[0043] The first calculation unit is used to obtain a peak distribution diagram corresponding to the binary image by summing the pixel values ​​of the binary image in the horizontal direction and assigning a pixel value less than a preset value to zero;

[0044] The second calculation unit is used to calculate the number of image pixels between the last pixel point of the first peak and the first pixel point of the last peak in the peak distribution graph;

[0045] The third calculation unit is used to obtain the thickness of the back fat based on the number of image pixels through a preset mapping function. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a schematic flow chart of an embodiment of a backfat thickness detection method provided by the present application;

[0047] Figure 2 It is a B-ultrasound image of the pig's back provided by this application;

[0048] Figure 3 It is the area image detected by the YOLO detection model provided in this application;

[0049] Figure 4 is the cropped backfat area image provided by the present application;

[0050] Figure 5 is a binary image provided by this application;

[0051] Figure 6 It is a peak distribution diagram of pixel values ​​after horizontal summation processing of the binary image provided by the present application;

[0052] Figure 7 It is a structural schematic diagram of an embodiment of a backfat thickness detection device provided in the present application. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0054] It should be understood that the step numbers used in this article are only for the convenience of description and are not intended to limit the order in which the steps are executed.

[0055] It should be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, the singular forms "a", "an" and "the" are intended to include plural forms unless the context clearly indicates otherwise.

[0056] The terms “include” and “comprising” indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0057] The term "and / or" means and includes any and all possible combinations of one or more of the associated listed items.

[0058] Backfat thickness refers to the thickness of subcutaneous fat at a specific location on the pig's back. It is an important indicator for assessing pork grade and a parameter that companies must measure during production. It can directly reflect the grade of pig carcasses. At present, domestic and foreign veterinary ultrasound instruments are mainly used to collect images, and then lines are drawn on the images and measured manually. However, this method relies on the operator's experience, has subjective errors, and poor accuracy. At the same time, manual measurement is time-consuming and inefficient, and it is difficult to meet large-scale needs.

[0059] Embodiment 1

[0060] Please refer to Figure 1 , Figure 1It is a flow chart of an embodiment of a backfat thickness detection method provided by the present application, including steps S1 to S4;

[0061] Step S1, obtaining a pig's back image;

[0062] It is understood that the back image is a B-ultrasound image of the pig's third to fourth ribs measured by a veterinary B-ultrasound instrument in a vertical method. The back image should completely and clearly include the back fat area. Figure 2 shown.

[0063] In this way, the B-ultrasound image of the pig's back is measured by a veterinary B-ultrasound instrument, which facilitates the subsequent accurate extraction of the back fat area.

[0064] Step S2, inputting the back image into a preset yolo detection model to extract a back fat area image in the back image;

[0065] It can be understood that after obtaining the back image of the pig, the back image can be input into the preset yolo detection model, and the eye muscle area and the back fat area in the back image can be identified and distinguished by the yolo detection model; by cropping the back fat area in the back image, the back fat area image in the back image can be obtained, wherein the area image detected by the yolo detection model is as follows Figure 3 As shown in the figure, the cropped back fat area image is as follows Figure 4 shown.

[0066] It should be noted that the yolo detection model can adopt the yolov9 algorithm, etc., and this application does not limit it.

[0067] In this way, the back fat area image in the back image can be accurately extracted through the yolo detection model, which facilitates the subsequent accurate calculation of the back fat thickness according to the number of pixels in the back fat area.

[0068] It should be noted that the training process of the yolo detection model is specifically as follows: a number of actual back ultrasound images of pigs are obtained through a veterinary ultrasound instrument, and blurred and unclear images are discarded; then the actual back fat areas in the actual back ultrasound images are labeled using the labelme software to obtain the labeling results; finally, the labeling results are input into the yolo detection model to train the yolo detection model.

[0069] In this way, by training the yolo detection model, the yolo detection model can be used to accurately extract the back fat area image in the back image.

[0070] Step S3, binarizing the back fat area image to obtain a binary image;

[0071] It is understandable that the backfat area image is first grayscale processed. Since each pixel in the grayscale image has only one brightness value, this value is usually between 0 (black) and 255 (white); the threshold is set to 185, and the pixels with grayscale values ​​less than the threshold 185 are set to 0, and the image with grayscale values ​​greater than or equal to the threshold 185 is set to 255, so that a binary image can be realized, which can simplify the image data and make image analysis easier, thereby better determining the location of the pig backfat fascia. Among them, the binary image is as follows: Figure 5 shown.

[0072] It should be noted that the above selection of the threshold in the binarization process is illustrative rather than restrictive, and those skilled in the art can set it according to actual needs.

[0073] Step S4, calculating the number of image pixels in the binary image corresponding to the backfat area, and obtaining the backfat thickness through a preset mapping function, wherein the mapping function is pre-established between the backfat thickness value and the number of image pixels.

[0074] It is understandable that after the back fat area image is converted into a binary image, the number of image pixels in the binary image can be calculated. Specifically, first, by summing the pixel values ​​of the binary image in the horizontal direction, the pixel sum of the binary image in the horizontal direction is obtained, and the pixel values ​​less than the preset value are assigned to zero, and the peak distribution diagram corresponding to the binary image is obtained, wherein the preset value can be set to two-thirds of the maximum value, that is, after calculating the pixel sum in the horizontal direction, the center position of the pig muscle fascia is more accurately located by setting the pixels less than two-thirds of the maximum pixel value to 0 (black). Secondly, calculate the number of image pixels between the last pixel point of the first peak and the first pixel point of the last peak in the peak distribution diagram to obtain the position difference between the first peak and the last peak; finally, based on the number of image pixels, the back fat thickness is obtained through a preset mapping function, wherein the preset mapping function is a linear function set in advance according to the number of image pixels. For example, when the back fat thickness corresponding to 55 pixels is 11.45 mm, k=4.8035 can be calculated through the linear function Y=kx. Multiple slopes k are calculated through multiple pixel numbers, and the slopes k are averaged to obtain the mapping function between the number of pixels and the back fat thickness. The peak distribution diagram of the pixel value after horizontal summation of the binary image is shown in FIG. Figure 6 shown.

[0075] In this way, by calculating the number of image pixels in the binary image corresponding to the backfat area, and calculating the backfat thickness through a mapping function pre-established between the backfat thickness value and the number of image pixels, the backfat thickness in the backfat area can be accurately detected, which has wide applicability.

[0076] The present application can accurately extract the back fat area image in the back image through the yolo detection model, so as to facilitate the subsequent accurate calculation of the back fat thickness according to the number of pixels in the back fat area; by binarizing the back fat area image, the center position of the pig's back fat fascia can be better determined, thereby accurately locating the pig's back fat pixels; by calculating the number of image pixels in the binarized image corresponding to the back fat area, and calculating the back fat thickness through a mapping function pre-established between the back fat thickness value and the number of image pixels, the back fat thickness in the back fat area can be accurately detected, and the application has wide applicability.

[0077] Please refer to Figure 7 , Figure 7 It is a structural schematic diagram of an embodiment of a backfat thickness detection device provided by the present application, comprising: an acquisition module 01, an extraction module 02, a processing module 03 and a calculation module 04;

[0078] The acquisition module 01 is used to acquire the back image of the pig;

[0079] It is understood that the back image is a B-ultrasound image of the pig's third to fourth ribs measured by a veterinary B-ultrasound instrument in a vertical method. The back image should completely and clearly include the back fat area. Figure 2 shown.

[0080] In this way, the B-ultrasound image of the pig's back is measured by a veterinary B-ultrasound instrument, which facilitates the subsequent accurate extraction of the back fat area.

[0081] The extraction module 02 is used to input the back image into a preset yolo detection model to extract the back fat area image in the back image;

[0082] The extraction module 02 includes: an input unit and a cropping unit; the input unit is used to input the back image into a preset yolo detection model to obtain the eye muscle area and back fat area in the back image; the cropping unit is used to obtain the back fat area image in the back image by cropping the back fat area in the back image.

[0083] It can be understood that after obtaining the back image of the pig, the back image can be input into the preset yolo detection model, and the eye muscle area and the back fat area in the back image can be identified and distinguished by the yolo detection model; by cropping the back fat area in the back image, the back fat area image in the back image can be obtained, wherein the area image detected by the yolo detection model is as follows Figure 3 As shown in the figure, the cropped back fat area image is as follows Figure 4 shown.

[0084] It should be noted that the yolo detection model can adopt the yolov9 algorithm, etc., and this application does not limit it.

[0085] In this way, the back fat area image in the back image can be accurately extracted through the yolo detection model, which facilitates the subsequent accurate calculation of the back fat thickness according to the number of pixels in the back fat area.

[0086] It should be noted that the training process of the yolo detection model is specifically as follows: a number of actual back ultrasound images of pigs are obtained through a veterinary ultrasound instrument, and blurred and unclear images are discarded; then the actual back fat areas in the actual back ultrasound images are labeled using the labelme software to obtain the labeling results; finally, the labeling results are input into the yolo detection model to train the yolo detection model.

[0087] In this way, by training the yolo detection model, the yolo detection model can be used to accurately extract the back fat area image in the back image.

[0088] The processing module 03 is used to perform binarization processing on the back fat area image to obtain a binary image;

[0089] It is understandable that the backfat area image is first grayscale processed. Since each pixel in the grayscale image has only one brightness value, this value is usually between 0 (black) and 255 (white); the threshold is set to 185, and the pixels with grayscale values ​​less than the threshold 185 are set to 0, and the image with grayscale values ​​greater than or equal to the threshold 185 is set to 255, so that a binary image can be realized, which can simplify the image data and make image analysis easier, thereby better determining the location of the pig backfat fascia. Among them, the binary image is as follows: Figure 5 shown.

[0090] It should be noted that the above selection of the threshold in the binarization process is illustrative rather than restrictive, and those skilled in the art can set it according to actual needs.

[0091] The calculation module 04 is used to calculate the number of image pixels in the binary image corresponding to the backfat area, and obtain the thickness of the backfat through a preset mapping function, wherein the mapping function is pre-established between the backfat thickness value and the number of image pixels.

[0092] The calculation module 04 includes: a first calculation unit, a second calculation unit and a third calculation unit; the first calculation unit is used to obtain a peak distribution diagram corresponding to the binary image by summing the pixel values ​​of the binary image in the horizontal direction and assigning pixel values ​​less than a preset value to zero; the second calculation unit is used to calculate the number of image pixels between the last pixel point of the first peak and the first pixel point of the last peak in the peak distribution diagram; the third calculation unit is used to obtain the thickness of the back fat based on the number of image pixels through a preset mapping function.

[0093] It is understandable that after the back fat area image is converted into a binary image, the number of image pixels in the binary image can be calculated. Specifically, first, by summing the pixel values ​​of the binary image in the horizontal direction, the pixel sum of the binary image in the horizontal direction is obtained, and the pixel values ​​less than the preset value are assigned to zero, and the peak distribution diagram corresponding to the binary image is obtained, wherein the preset value can be set to two-thirds of the maximum value, that is, after calculating the pixel sum in the horizontal direction, the center position of the pig muscle fascia is more accurately located by setting the pixels less than two-thirds of the maximum pixel value to 0 (black). Secondly, calculate the number of image pixels between the last pixel point of the first peak and the first pixel point of the last peak in the peak distribution diagram to obtain the position difference between the first peak and the last peak; finally, based on the number of image pixels, the thickness of the back fat is obtained through a preset mapping function, wherein the preset mapping function is a linear function that is set in advance according to the pixels. For example, when the back fat thickness corresponding to 55 pixels is 11.45 mm, k=4.8035 can be calculated through the linear function Y=kx. Multiple slopes k are calculated through multiple pixel numbers, and the slopes k are averaged to obtain the mapping function between the number of pixels and the back fat thickness. The peak distribution diagram of the pixel value after horizontal summation of the binary image is shown in FIG. Figure 6 shown.

[0094] In this way, by calculating the number of image pixels in the binary image corresponding to the backfat area, and calculating the backfat thickness through a mapping function pre-established between the backfat thickness value and the number of image pixels, the backfat thickness in the backfat area can be accurately detected, which has wide applicability.

[0095] The present application can accurately extract the back fat area image in the back image through the yolo detection model, so as to facilitate the subsequent accurate calculation of the back fat thickness according to the number of pixels in the back fat area; by binarizing the back fat area image, the center position of the pig's back fat fascia can be better determined, thereby accurately locating the pig's back fat pixels; by calculating the number of image pixels in the binarized image corresponding to the back fat area, and calculating the back fat thickness through a mapping function pre-established between the back fat thickness value and the number of image pixels, the back fat thickness in the back fat area can be accurately detected, and the application has wide applicability.

[0096] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and is not intended to limit the scope of protection of the present application. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of protection of the present application.

Claims

1. A method for detecting backfat thickness, characterized in that: include: Obtain an image of the pig's back; Inputting the back image into a preset yolo detection model to extract a back fat area image in the back image; Binarizing the backfat region image to obtain a binary image; The number of image pixels in the binary image corresponding to the backfat area is calculated, and the thickness of the backfat is obtained through a preset mapping function, wherein the mapping function is pre-established between the backfat thickness value and the number of image pixels.

2. A backfat thickness detection method according to claim 1, characterized in that: The back image is a B-ultrasound image measured by a veterinary B-ultrasound instrument.

3. A backfat thickness detection method according to claim 1, characterized in that: The back image is input into a preset yolo detection model to extract the back fat area image in the back image, specifically: Input the back image into a preset yolo detection model to obtain the eye muscle area and back fat area in the back image; The back fat region image in the back image is obtained by cropping the back fat region in the back image.

4. The backfat thickness detection method according to claim 1, characterized in that: The training process of the YOLO detection model is as follows: Obtain several actual back B-ultrasound images of pigs; Using labelme to label actual back fat areas in a plurality of actual back B-ultrasound images to obtain labeling results; The annotation results are input into the YOLO detection model to train the YOLO detection model.

5. The backfat thickness detection method according to claim 1, characterized in that: The number of image pixels in the binary image corresponding to the backfat area is calculated, and the thickness of the backfat is obtained by a preset mapping function, specifically: By summing the pixel values ​​of the binary image in the horizontal direction and assigning a pixel value less than a preset value to zero, a peak distribution diagram corresponding to the binary image is obtained; Calculate the number of image pixels between the last pixel point of the first peak and the first pixel point of the last peak in the peak distribution graph; Based on the number of image pixels, the thickness of the back fat is obtained through a preset mapping function.

6. A backfat thickness detection device, characterized in that: include: Acquisition module, extraction module, processing module and calculation module; The acquisition module is used to acquire the back image of the pig; The extraction module is used to input the back image into a preset yolo detection model to extract a back fat area image in the back image; The processing module is used to perform binarization processing on the back fat area image to obtain a binary image; The calculation module is used to calculate the number of image pixels in the binary image corresponding to the backfat area, and obtain the thickness of the backfat through a preset mapping function, wherein the mapping function is pre-established between the backfat thickness value and the number of image pixels.

7. The backfat thickness detection device according to claim 6, characterized in that: The back image is a B-ultrasound image measured by a veterinary B-ultrasound instrument.

8. The backfat thickness detection device according to claim 6, characterized in that: The extraction module includes: an input unit and a cutting unit; The input unit is used to input the back image into a preset YOLO detection model to obtain the eye muscle area and back fat area in the back image; The cropping unit is used to obtain the back fat area image in the back image by cropping the back fat area in the back image.

9. The backfat thickness detection device according to claim 6, characterized in that: The training process of the YOLO detection model is as follows: Obtain several actual back B-ultrasound images of pigs; Using labelme to label actual back fat areas in a plurality of actual back B-ultrasound images to obtain labeling results; The annotation results are input into the YOLO detection model to train the YOLO detection model.

10. The backfat thickness detection device according to claim 6, characterized in that: The calculation module includes: a first calculation unit, a second calculation unit and a third calculation unit; The first calculation unit is used to obtain a peak distribution diagram corresponding to the binary image by summing the pixel values ​​of the binary image in the horizontal direction and assigning a pixel value less than a preset value to zero; The second calculation unit is used to calculate the number of image pixels between the last pixel point of the first peak and the first pixel point of the last peak in the peak distribution graph; The third calculation unit is used to obtain the thickness of the back fat based on the number of image pixels through a preset mapping function.

Images