Method for positioning cutting position of meat loaf and meat loaf cutting equipment

Abstract

The invention discloses a method for positioning the cutting position of meat loaves and equipment for cutting the meat loaves. The positioning method comprises the steps of obtaining an original image of an end face of a meat block; identifying the original image to obtain an identification image including a target part; and selecting a segmentation position according to the target part. The meat block cutting equipment comprises a detection device, an industrial control computer, an electrical control device, a cutting device and a control motor. According to the method and equipment disclosed by the invention, the convenience and diversity of meat block cutting can be improved, and the use of manpower is reduced.

Description

Technical Field

[0001] The present disclosure generally relates to the technical field of meat block segmentation. More specifically, the present disclosure relates to a method for positioning the segmentation position of a meat block and a meat block segmentation device. Background Art

[0002] In a pork processing factory, the quantity of meat block products containing backbone fat is large. It is necessary to separate the backbone fat part from other parts of the meat block and process them into different meat products for sale, and it is impossible to rely entirely on manual labor to complete.

[0003] Existing backbone fat segmentation devices can achieve semi-automatic segmentation. The device includes an adjustable tool holder. When performing the segmentation action, the height of the tool holder can be adjusted according to characteristics such as the thickness of the backbone fat, and the start and stop of the device are controlled by manually operating a foot pedal switch. However, during the segmentation of the meat block, the height cannot be adjusted in real time according to the characteristics of the backbone fat thickness of the meat block. It is necessary to stop the device, adjust the height of the tool holder to a suitable position by the staff, and then restart the device for segmentation. Moreover, for the segmentation of the meat block, the confirmation of the segmentation position can only rely on manual labor at present, with low efficiency, and there is no corresponding automated meat block segmentation device on the market.

[0004] In view of this, there is an urgent need to provide a method for positioning the segmentation position of a meat block and a meat block segmentation device, so as to improve the convenience and diversity of meat block segmentation and reduce the use of manual labor. Summary of the Invention

[0005] In order to solve at least one or more of the above-mentioned technical problems, the present disclosure provides, in multiple aspects, a method for positioning the segmentation position of a meat block and a meat block segmentation device, which can realize the positioning of the segmentation positions of various meat blocks to achieve the automated segmentation of meat blocks.

[0006] In a first aspect, the present disclosure provides a method for positioning the segmentation position of a meat block, including: obtaining an original image of an end face of the meat block; identifying the original image to obtain an identification image including a target part; and selecting a segmentation position according to the target part.

[0007] In a second aspect, the present disclosure provides an apparatus for cutting meat blocks, characterized by comprising: a detection device, an industrial control computer, an electrical control device, a cutting device, and a control motor; wherein, the detection device is configured to acquire original image data of the end face of a meat block; the industrial control computer is configured to execute the cutting position positioning method in the first aspect and transmit the obtained target part information to the electrical control device; the electrical control device adjusts the height of the cutting device according to the target part information to ensure that the cutting device is aligned with the target part; the cutting device is configured to perform a cutting action to cut the meat block; and the control motor is configured to receive an instruction from the electrical control device and dynamically adjust the height of the cutting device.

[0008] Through a method for positioning the cutting position of a meat block and a meat block cutting apparatus provided as above, the embodiments of the present disclosure acquire an original image of the end face of a meat block; identify the original image to obtain an identification image including a target part, and can select a cutting position according to the target part. In some embodiments, the original image of the end face of a meat block containing back fat is identified to obtain an identification image with the back fat part as the target part.

[0009] Further, in some embodiments, the thickness of the target part is determined, a central region with a specific width is selected in the target part; the central region is divided into a plurality of sub-regions; some sub-regions are selected as sampled sub-regions, and the thickness of each sub-region in the sampled sub-regions is calculated; the weighted average value of the thicknesses of the sampled sub-regions is calculated to determine the thickness of the target part, and thus the cutting position is determined. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present disclosure will become readily understood. In the drawings, several embodiments of the present disclosure are shown in an exemplary rather than restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0011] Figure 1 shows the original image of the end face of the meat block according to the embodiment of the present disclosure;

[0012] Figure 2A shows a schematic diagram of the hardware composition of a meat block cutting apparatus according to the embodiment of the present disclosure;

[0013] Figure 2B shows a top view of the movement of the camera 212 when the position of the meat block on the conveyor belt is not correct according to the embodiment of the present disclosure;

[0014] Figure 3Shows an exemplary flowchart of a method 300 for positioning the cutting positions of meat blocks in the present disclosure;

[0015] Figure 4 Shows a schematic diagram of the placement of a meat block in an embodiment of the present disclosure;

[0016] Figure 5 Shows that in an embodiment of the present disclosure, after semantic segmentation of the Figure 1 original image according to the U-net semantic segmentation network;

[0017] Figure 6 Shows a processing schematic diagram of performing opening and closing operations on the irregular regions in the backfat region image in an embodiment of the present disclosure;

[0018] Figure 7 Shows an exemplary flowchart for determining the thickness of a target part in an embodiment of the present disclosure;

[0019] Figure 8 Shows a schematic diagram of selecting a central region with a specific width in a target part in an embodiment of the present disclosure;

[0020] Figure 9 Shows an exemplary flowchart for calculating the weighted average of the thicknesses of the sampling sub-regions in an embodiment of the present disclosure;

[0021] Figure 10 Shows an identification image of the thickness of the backfat part in an embodiment of the present disclosure;

[0022] Figure 11 Shows an exemplary flowchart for confirming the cutting positions of a meat block including an intermuscular fat part in an embodiment of the present disclosure;

[0023] Figure 12 Shows a schematic diagram of determining the cutting positions in the intermuscular fat part in an embodiment of the present disclosure. Detailed implementation manners

[0024] Next, the technical solutions in the embodiments of the present disclosure will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present disclosure. Obviously, the described embodiments are some, but not all, of the embodiments of the present disclosure. All other embodiments obtained by those skilled in the art based on the embodiments of the present disclosure without creative efforts shall fall within the protection scope of the present disclosure.

[0025] It should be understood that the terms "including" and "comprising" used in the specification and claims of the present disclosure indicate the presence of the described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their combinations.

[0026] It should also be understood that the terms used in this disclosure specification are merely for the purpose of describing specific embodiments and are not intended to limit the disclosure. As used in this disclosure specification and the claims, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms. It should be further understood that the term "and / or" used in this disclosure specification and the claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0027] As used in this specification and the claims, the term "if" can be interpreted as "when", "once", "in response to determining", or "in response to detecting" depending on the context. Similarly, the phrase "if determined" or "if [the described condition or event] is detected" can be interpreted as meaning "once determined", "in response to determining", "once [the described condition or event] is detected", or "in response to detecting [the described condition or event]" depending on the context.

[0028] The specific embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0029] Figure 1 The original image of the end face of the meat block in the embodiment of the present disclosure is shown. In the figure, the meat block is the meat block after preliminary processing after the pig is slaughtered. The meat block includes three parts: other parts 110, tenderloin part 120, and back fat part 130. Among them, the other parts 110 are some minced meat and bones in the meat block. The tenderloin part 120 is the strip-shaped lean meat inside the pig's spine bone. The back fat part 130 is the whole piece of fat without red meat on the outside of the pig's back spine bone. The fat in the back fat part generally does not contain lymph, has a higher oil yield than the fat in other parts, and has a wider applicability. Therefore, the back fat part needs to be separated from other different parts and sold or further processed as different meat products. Since the back fat part is adjacent to the pig skin, generally the back fat part and the pig skin are separated together as a whole and combined into the product of back fat skin.

[0030] The embodiment of the present disclosure provides a meat block splitting device. By acquiring the original image of the end face of the meat block, the device can identify the original image of the end face of the meat block to obtain an identification image including the target part, and select the splitting position according to the target part to automatically split the meat block.

[0031] Figure 2A The schematic diagram of the hardware composition of a meat block splitting device in the embodiment of the present disclosure is shown. As shown in the figure, the device includes a detection device 210, a control computer 220, an electrical control device 230, a control motor 240, and a splitting device 250. Among them, the detection device 210 includes an infrared sensing device 211 and a camera 212.

[0032] The detection device 201 is used to detect the meat blocks conveyed on a conveyor belt (not shown) and capture the original image of the end face of the meat blocks. Specifically, when the infrared sensing device 211 senses the arrival of the meat blocks on the conveyor belt, it transmits the arrival information of the meat blocks to the control computer 220, and then the control computer 220 sends an instruction to the camera 212, so that the camera 212 performs a shooting action on the end face of the meat blocks and transmits the captured original image of the end face of the meat blocks back to the control computer 220.

[0033] In one embodiment, the arrangement of the camera 212 is movable. When the position of the meat block is not correct relative to the initial position of the camera 212, the control computer 220 can analyze the position of the meat block on the conveyor belt through the laser ranging devices arranged on both sides of the conveyor belt, so as to control the movement of the camera 212 to the position directly facing the end face of the meat block, so that the camera 212 can capture a clear and correctly positioned original image of the end face of the meat block.

[0034] Figure 2B The top view shows the movement of the camera 212 when the position of the meat block on the conveyor belt is not correct in one embodiment. As Figure 2B shown, the conveyor belt conveys the meat block 260 in front of the camera 212, and laser ranging devices 271 and 272 are arranged on both sides of the conveyor belt. The laser ranging device 271 and the laser ranging device 272 respectively measure the distances to the meat block 260 and transmit them to the control computer 220. The control computer 220 calculates the distance difference between the meat block 260 and the laser ranging device 271 and the laser ranging device 272, so as to adjust the moving distance and moving direction of the camera. For example, the distance between the meat block 260 and the laser ranging device 271 is 70 cm, and the distance between the meat block 260 and the laser ranging device 272 is 90 cm. At this time, the distance difference between the meat block 260 and the laser ranging device 271 and the laser ranging device 272 is 20 cm, and the meat block 260 is closer to the laser ranging device 271. Then the control computer 220 controls the camera 212 to move 20 cm in the direction of the laser ranging device 271 to reach the front of the meat block 260 to capture the original image of the end face of the meat block 260.

[0035] In one embodiment, the camera 212 in the detection device 201 can be an industrial camera. Compared with the image sensor of an ordinary camera that performs interlaced scanning, the image sensor of an industrial camera performs progressive scanning, and the imaging is more accurate. And the industrial camera has a compact and strong structure and is not easy to be damaged. For the complex environment on the production line in a meat processing plant, it can minimize the influence and keep the camera working properly for a long time.

[0036] The control computer 220 is used to control the infrared induction device 211 to sense the meat blocks arriving on the conveyor belt, issue an instruction to the camera 212 to capture the original image of the end face of the meat block, analyze and identify the original image of the end face of the meat block, then find a suitable cutting position, and transmit the feature information including the cutting position of the meat block to the electrical control device.

[0037] The electrical control device 230 is used to receive the information from the control computer 220, which is the feature information including the cutting position of the meat block analyzed and identified by the control computer 220 based on the original image of the end face of the meat block. Based on the feature information including the cutting position of the meat block, the electrical control device 230 issues an instruction to the lifting device 240 to adjust the height of the cutting device 250 so that the cutting device 250 cuts the meat block at the identified cutting position. In the embodiment of the present disclosure, the electrical control device 230 is a PLC (Programmable Logic Controller).

[0038] The lifting device 240 is used to receive the instruction from the electrical control device 230 and adjust the height of the cutting device 250 in real time. When the cutting device finishes processing the previous meat block and needs to process other subsequent meat blocks on the conveyor belt, the backfat thickness or the distribution position of each meat block may not be exactly the same, that is, the cutting position of each meat block may not be the same. Therefore, each time a new meat block is cut, the height of the cutting device 250 needs to be adjusted to adapt to the backfat thickness of different meat blocks. For example, when the cutting position calculated by the control computer 220 drops by 2 cm relative to the previous meat block, the lifting device 240 correspondingly adjusts the cutting device 250 to drop by 2 cm; when the cutting position calculated by the control computer 220 rises by 2 cm, the lifting device 240 correspondingly adjusts the cutting device 250 to rise by 2 cm. By adjusting the height of the cutting device 250 in real time through the lifting device 240, the labor is saved and the cutting efficiency is improved at the same time.

[0039] The above describes a meat block cutting device provided by the embodiment of the present disclosure. Correspondingly, the embodiment of the present disclosure also provides a method for positioning the cutting position of a meat block, including: at step S310, obtaining the original image of the end face of the meat block; at step S320, identifying the original image to obtain an identification image including the target part; at step S330, selecting a cutting position according to the target part. Figure 3 An exemplary flowchart of a method 300 for positioning the cutting position of a meat block according to the embodiment of the present disclosure is shown.

[0040] Specifically, at step S310, an original image of the end face of the meat block is obtained. The way to obtain the image is to convey the meat block to the front of the meat block splitting device through a conveyor belt. A detection device is arranged in front of the meat block splitting device to detect the end face of the meat block conveyed on the conveyor belt. The specific detection includes using an infrared induction device to sense the arrival of the meat block. The infrared induction device transmits the sensed arrival information of the meat block to the control computer. Subsequently, the control computer issues an instruction to the camera to capture the end face of the meat block and transmit the image to the control computer, thereby obtaining the original image of the end face of the meat block.

[0041] The preliminarily processed meat block does not have a standard shape, but generally conforms to an approximate shape of a cuboid. Figure 4 FIG. shows a schematic diagram of the placement of a meat block according to an embodiment of the present disclosure. Among them, the lower bottom surface is the pigskin of the meat block. The meat block is generally placed on the conveyor belt with the side with pigskin facing down. The side end cross-sections 410, 420, 430, and 440 of the meat block are the end faces of the meat block. Therefore, there are 4 end faces of the meat block, which are divided into two categories. One is the two end faces 430 and 440 on the longer side of the length, and the other is the two end faces 410 and 420 on the shorter side of the length. In the embodiment of the present disclosure, the original image of the end face on the shorter side of the length of the meat block is obtained.

[0042] Next, at step S320, the original image of the end face of the obtained meat block is recognized to obtain a recognition image including the target part. In order to obtain the specific splitting position, it is necessary to first recognize the target part including the splitting position. By performing semantic segmentation on the original image of the end face of the meat block, a mask image including different regions is obtained, and then the recognition image including the target part is extracted. In the embodiment of the present disclosure, through the U-net semantic segmentation network, the original image of the end face of the meat block is segmented into different regions to obtain a mask image including different regions.

[0043] Figure 5 FIG. shows the schematic diagram after semantic segmentation of the Figure 1 original image according to the U-net semantic segmentation network in the embodiment of the present disclosure. The original image of the end face of the meat block is input into the U-net semantic segmentation network, and a mask image including three regions 510, 520, and 530 as shown in Figure 5 is segmented. Among them, 510 is the image of other regions of the meat block, including other minced meat parts or bones. 520 is the image of the tenderloin region, including the tenderloin part. 530 is the image of the back fat region, including the back fat part and the fine minced part. In the embodiment of the present disclosure, the back fat region image 530 is the recognition image including the target part.

[0044] In one embodiment, when it is necessary to cut the tenderloin part or other parts of the meat block, the tenderloin region image or other region images can also be used as the recognition image of the target part. Those skilled in the art can understand that the embodiments of the present disclosure have no limitations in this regard.

[0045] In yet another embodiment, for semantic segmentation of the original image of the end face of the meat block, other semantic segmentation networks can also be used. For example, semantic segmentation of the original image of the end face of the meat block can be performed through semantic segmentation networks such as FCN, SegNet, and PSPNet. Those skilled in the art can understand that the embodiments of the present disclosure have no limitations in this regard.

[0046] Finally, at step S330, the cutting position is determined according to the recognition image including the target part. In the embodiments of the present disclosure, the recognition images including different regions after semantic segmentation are separately extracted to obtain the binary recognition images of each region, which is more conducive to subsequent recognition and sampling. For example, the mask images and the original images of different regions obtained after semantic segmentation are subjected to bitwise AND operation based on pixel points, so as to extract the binary recognition images of each region.

[0047] When the original image of the end face of the meat block is taken, some other things in the surrounding environment may be taken into the original image, resulting in some noise points in the image or unclosed holes. Although the semantic segmentation network segments the original image, the irregular regions still exist. The irregular regions can be removed by means of image opening operation and image closing operation. Taking the backfat region as an example, Figure 6 FIG. shows a schematic diagram of the processing of the irregular regions in the backfat region image in the embodiments of the present disclosure. The backfat region image in 610 contains noise points misrecognized from the surrounding environment. The noise points in 610 are removed through image opening operation to obtain the backfat region image in 620. The backfat region image in 630 has some internal small holes. The small holes in 630 are closed through image closing operation to obtain the backfat region image in 640.

[0048] In order to cut the backfat part from the meat block, the selection of the cutting position can be determined according to the internal boundary points of the backfat part. In the embodiments of the present disclosure, positioning the cutting position according to the target part includes: determining the thickness of the target part; and positioning the cutting position according to the thickness of the target part. That is, it is necessary to determine the thickness of the backfat part, and the boundary point coordinates close to the inside of the meat block in the thickness data are used as the specific cutting position.

[0049] The embodiments of the present disclosure provide a method 700 for determining the thickness of a target part, including: at step S710, selecting a central region with a specific width in the target part; at step S720, dividing the central region into a plurality of sub-regions; at step S730, selecting some of the sub-regions as sampled sub-regions and calculating the thickness of each sub-region in the sampled sub-regions; at step S740, calculating the weighted average of the thicknesses of the sampled sub-regions to determine the thickness of the target part. Figure 7 An exemplary flowchart for determining the thickness of a target part is shown.

[0050] Specifically, at step S710, a central region with a specific width is selected in the target part. In the embodiments of the present disclosure, based on the identified image of the backfat region described above, a central region of the image is selected for subsequent sampling, and the selected image width is 50 pixel widths. Figure 8 A schematic diagram showing the selection of a central region with a specific width in the target part is shown. Among them, image 810 is the binary image of the backfat region. A central region with a width of 50 pixel widths is selected in image 810 to obtain central region 820. In addition, when the target part is included in the tenderloin region, a central region with a width of 50 pixel widths can be selected in the binary image 830 of the tenderloin region to obtain central region 840.

[0051] Next, at step S720, the selected central region is divided into a plurality of sub-regions. Since the backfat thickness of the selected central region is not necessarily uniform, and the cutting point must be fixed, in order to cut off as much of the backfat part in the meat block as possible, it is necessary to understand the thickness weight in the selected central region. Therefore, the selected central region is divided into a plurality of sub-regions. In the embodiments of the present disclosure, each sub-region is evenly divided and spaced by a specific width. Among them, evenly dividing a plurality of sub-regions so that the pixel width of each sub-region is the same can reduce the calculation amount in the subsequent thickness statistics of multiple sub-regions; spacing by a specific width means that, along the thickness direction of the backfat part, an image every 3 pixel widths is used as a sub-region.

[0052] At step S730, some of the sub-regions are selected as sampled sub-regions, and the thickness of each sub-region in the sampled sub-regions is calculated. In the embodiments of the present disclosure, the multiple sub-regions divided in the previous step are sampled at intervals. By traversing the central region divided into multiple sub-regions, every other sub-region, the pixel values of this sub-region are set to zero to obtain a sampled sub-region containing some sub-regions. Among them, calculating the thickness of each sub-region in the sampled sub-region is to calculate the minimum bounding rectangle of each sub-region in the sampled sub-region and count the lengths of these minimum bounding rectangles, and use the length of the minimum bounding rectangle as the thickness of each sub-region in the sampled sub-region.

[0053] At step S740, a weighted average of the thicknesses of the sampled sub-regions is calculated to determine the thickness of the target site. Since the sampled sub-regions contain sub-regions with various thicknesses, in order to make the calculated thickness of the target site closer to the thicknesses of most sub-regions, it is necessary to calculate the weighted average of the thicknesses of the sampled sub-regions.

[0054] In an embodiment of the present disclosure, a specific calculation method 900 for calculating the weighted average of thickness data is given. Figure 9 An exemplary flowchart for calculating the weighted average of the thicknesses of the sampled sub-regions is shown.

[0055] At step S910, the data distribution variance of the thicknesses of the sampled sub-regions is calculated. Variance is a representative descriptive value used to reflect the degree of fluctuation within a set of data. Calculating the data distribution variance of the thicknesses of the sampled sub-regions can reflect the distribution of different thickness data in the sampled sub-regions.

[0056] At step S920, based on the data distribution variance, the sub-region thickness data with the top N frequencies of occurrence are selected. In an embodiment of the present disclosure, a corresponding regression network is designed based on a BP neural network. The variance value of the thickness data of the sampled sub-regions is input into the regression network, and in the regression network, the input is mapped to the output, and the obtained output is the number of thickness data to be included in the weighted average calculation, that is, the sub-region thickness data with the top N frequencies of occurrence. Among them, in the thickness data with the top N frequencies of occurrence, N is the number of types of thickness data selected by the regression network based on the BP neural network.

[0057] At step S1030, the weighted average of the thicknesses of the partial sub-regions is calculated according to the frequencies of occurrence and the thickness data of the sub-regions with the top N frequencies of occurrence. In an embodiment of the present disclosure, the weighted average is calculated using the following formula:

[0058]

[0059]

[0060] Among them, n represents the number of times each thickness data appears, N is the total number of types of thickness data included in the weighted average formula, the value of i ranges from 1 to F, F is the optimal total number of data to be included in the weighted average obtained according to the regression network, r is the thickness data under this weight included in the weighted average, and R is the weighted average of the sub-region thicknesses with the top N frequencies of occurrence. Through the above formula, the weighted average of the thicknesses of the sampled sub-regions can be obtained.

[0061] After determining the thickness data of the target site, the boundary point coordinates of the thickness are used as the segmentation position. Figure 10An identification image of the backfat thickness of the embodiment of the present disclosure is shown. In the figure, points A and B are the thickness boundary points obtained according to the above thickness calculation method 900. Point A is the upper boundary point of the backfat part, and point B is the lower boundary point of the backfat part. Among them, point A is located on the side close to the tenderloin part, and point B is located on the side of the meat block with skin. In the embodiment of the present disclosure, the upper boundary point of the backfat part is used as the segmentation position, that is, point A is used as the determined segmentation position. To prevent the segmentation device from performing a non-horizontal segmentation during the segmentation process but instead cutting through the backfat part downward, the lower boundary point B of the thickness is used as the guaranteed limit of the segmentation position. When the segmentation device is too low and exceeds the lower boundary point B of the backfat part, the segmentation device is adjusted back to the initial height.

[0062] In another embodiment of the present disclosure, the meat block to be segmented further includes intermuscular fat. The intermuscular fat exists in the minced part of the meat block, and the minced part is located in the middle of the backfat part and the tenderloin part, and includes red meat and intermuscular fat. To make the segmented tenderloin part more pure, the minced part and the backfat part can be segmented together, and the target part at this time is the intermuscular fat in the minced part. Based on the foregoing method 300, at step 320, the U-net semantic segmentation network is also used to perform semantic segmentation on the original image of the meat block including the intermuscular fat part, and identification images of the backfat region, the tenderloin region, and other regions are obtained. At this time, the backfat region image includes the intermuscular fat part and the backfat part.

[0063] In the embodiment of the present disclosure, a method 1100 for confirming the segmentation position of the intermuscular fat part is given. Figure 11 An exemplary flowchart for confirming the segmentation position of the meat block including the intermuscular fat part in the embodiment of the present disclosure is shown. At step S1110, a central region with a specific width is selected in the target part. The selection method is the same as the method in step 710 of the foregoing method 700. Based on the identification image of the backfat region including the intermuscular fat, a central region with a width of 50 pixel widths is selected in the intermuscular fat part of the image.

[0064] At step S1120, the minimum bounding rectangle of the central region is calculated. In the embodiment of the present disclosure, calculating the minimum bounding rectangle is to enclose the central region selected in the previous step to form a rectangle parallel to the X-axis and the Y-axis.

[0065] Next, at step S1130, according to the minimum bounding rectangle, the center point of the target part is determined. In the embodiment of the present disclosure, based on the minimum bounding rectangle determined in the previous step, the center point of the rectangle is calculated, and the center point of the rectangle is used as the segmentation position. Figure 12A schematic diagram showing the determination of the segmentation position in the intermuscular fat part in an embodiment of the present disclosure is shown. Among them, the image 1210 is the central region of the intermuscular fat part obtained according to the foregoing method 700. Based on the image 1210, the minimum bounding rectangle is determined and the center point of this rectangle is calculated to obtain the point C in the image 1220, and the point C is the determined segmentation position.

[0066] Since the sizes of different parts of each piece of meat are not exactly the same, when the intermuscular fat part is too small to be recognized, the target part can be assisted in recognition according to the part adjacent to the intermuscular fat part. In the embodiment of the present disclosure, the tenderloin part adjacent to the intermuscular fat part is used to assist in recognizing the target part. The specific recognition method is to identify the boundary point in the tenderloin part in the direction close to the intermuscular fat, and select a point at a specific distance in the direction from this boundary point to the backfat part as the segmentation position.

[0067] In one embodiment, after determining the segmentation position in the intermuscular fat part, the thickness of the backfat part can be recognized according to the foregoing method 300, and the thickness of the backfat part is used as the bottom line of the segmentation position, so that during the process of dividing the piece of meat by the dividing device, it can be divided along the thickness of the backfat part to prevent uneven division.

[0068] Although multiple embodiments of the present disclosure have been shown and described herein, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art can think of many changes, alterations, and alternative ways without departing from the spirit and scope of the present disclosure. It should be understood that various alternative solutions to the embodiments of the present disclosure described herein can be adopted in the practice of the present disclosure. The appended claims are intended to define the scope of protection of the present disclosure and thus cover equivalents or alternatives within the scope of these claims.

Claims

1. A method for locating a cutting position of a meat block, comprising: Obtaining the original image of the end surface of the meat block; Recognize the original image to obtain a recognized image including the target part; as well as A segmentation position is selected according to the target part.

2. The method according to claim 1, characterized in that The meat block is a meat block containing spinal fat, and locating the cutting position according to the target part includes: determining the thickness of the target site; and The splitting position is located according to the thickness.

3. The method according to claim 2, characterized in that Determining the thickness of the target site includes: selecting a central region having a specific width in the target site; dividing the central area into a plurality of sub-areas; Selecting some sub-regions as sampling sub-regions, and calculating the thickness of each sub-region in the sampling sub-regions; A weighted average of the thicknesses of the sampling sub-areas is calculated to determine the thickness of the target site.

4. The method according to claim 3, characterized in that Select some sub-areas as sampling sub-areas including: Every time a sub-region is separated, the next sub-region is selected, and the sub-regions selected at these intervals are used as sampling sub-regions.

5. The method according to claim 3, characterized in that: Calculating the weighted average of the thickness of the sampling sub-areas includes: Calculate the data distribution variance of the thickness of the sampling sub-area; Based on the data distribution variance, selecting the thickness data of the sub-regions with the top N occurrence frequencies; The weighted average of the thicknesses of the sub-regions is calculated based on the occurrence frequencies and thickness data of the sub-regions with the top N occurrence frequencies.

6. The method according to claim 1, characterized in that The meat block is a meat block containing interlayer fat and spinal fat, and locating the cutting position according to the target part includes: determining a center point of the target site; and The segmentation position is located according to the center point.

7. The method according to claim 6, characterized in that Determining the center point of the target part includes: selecting a central region having a specific width in the target site; Calculate the minimum enclosing rectangle of the central area; The center point of the target part is determined according to the minimum enclosing rectangle.

8. The method according to claim 7, characterized in that When the interlayer fat part cannot be identified, the target part is assisted in identifying according to the part adjacent to the interlayer fat part.

9. The method according to claim 8, characterized in that Target areas to be identified based on the areas immediately adjacent to the intercalated fat area include: Identifying a boundary point in a region adjacent to the interlayer fat region and close to the interlayer fat region; Based on the boundary point close to the interlayer fat direction, a point at a specific distance in the direction of the spinal fat is selected as the target part.

10. A meat cutting device, characterized in that: include: Detection device, industrial control computer, electrical control device, splitting device and control motor; among which, The detection device is used to obtain original image data of the end surface of the meat block; The industrial control computer is used to execute the segmentation position positioning method described in claims 1 to 16, and transmit the obtained target part information to the electrical control device; The electrical control device adjusts the height of the segmentation device according to the target part information to ensure that the segmentation device is aligned with the target part; The dividing device is used to perform a dividing action to divide the meat block; The control motor is used to receive instructions from the electrical control device and dynamically adjust the height of the dividing device.

Images