Soft middle front segment segmentation method, system, device, equipment and medium

Abstract

The invention belongs to the technical field of slaughter processing equipment, and particularly discloses a soft middle front segment segmentation method, system, device and equipment and a medium, and the method comprises the steps: deploying an image collection device on a meat product segmentation production line, and obtaining an original image of a to-be-segmented soft middle front segment; inputting the original image into a pre-trained semantic segmentation model for reasoning analysis, accurately predicting the position of a crescent bone in the soft middle front segment, and generating a semantic segmentation image; based on the analysis of the semantic segmentation image, planning segmentation track points; and executing a segmentation scheme according to the segmentation track points. The method is based on semantic segmentation image analysis, the position of the crescent bone in the soft middle front segment is accurately predicted, the problem caused by manual labeling is solved, the segmentation precision and efficiency of the soft middle front segment are effectively improved, and the method is suitable for the requirements of modern meat product segmentation production lines.

Description

Technical Field

[0001] This invention belongs to the technical field of slaughtering and processing equipment, and in particular relates to a method, system, device, equipment and medium for soft mid-front segmentation. Background Technology

[0002] In the process of processing pork ribs, the soft mid-front section after separation of ribs needs to be manually located at the edge of the crescent bone to achieve precise division, separating the soft mid-front section into the soft mid section and the front section without ribs. Then, the soft mid section undergoes secondary processing and is cut into pork belly with skin and No. 3 meat products with fat.

[0003] Currently, the separation of the soft mid-front segment often relies on manual positioning of the crescent bone, which has a large identification error, affecting the yield of multi-colored products and also carries the risk of cutting the crescent bone, resulting in profit loss. At the same time, the processing of the soft mid-front segment produces a large number of products, and manual processing and separation of these products is time-consuming, labor-intensive, and inefficient.

[0004] Against this backdrop, designing a soft mid-front segmentation method, system, device, equipment, and medium to automatically identify and locate the crescent bone position through an algorithm model, replacing manual positioning, has become a key research topic in the existing technology field. Summary of the Invention

[0005] In order to overcome the above-mentioned shortcomings of the prior art, the present invention provides a method, system, device, equipment and medium for soft front-end segmentation, which solves the above problems.

[0006] To achieve the above objectives, the first aspect of the present invention discloses a soft mid-front segmentation method, the method comprising: Image acquisition devices are deployed on the meat processing production line to acquire raw images of the front section of the meat to be processed; The original image is input into a pre-trained semantic segmentation model for inference analysis, accurately predicting the position of the lunula inside the anterior segment of the soft midline, and generating a semantic segmentation image. Based on the analysis of semantic segmentation images, segmentation trajectory points are planned; The segmentation scheme is executed based on the segmentation trajectory points.

[0007] Compared with existing technologies, this application has the following advantages: This invention completely replaces manual marking of the crescent bone in the anterior segment of the soft midsection, significantly improving the accuracy and efficiency of crescent bone position marking on meat processing production lines, and adapting to the production needs of modern production lines. By using a semantic segmentation model for identification and marking, the influence of subjective human factors is eliminated, ensuring accurate marking of the crescent bone in the anterior segment of the soft midsection and preventing the risk of accidental damage to the crescent. Simultaneously, it innovatively utilizes a semantic segmentation model to first predict the crescent bone position, and then plans semantic segmentation trajectory points based on semantic segmentation image analysis to form a segmentation scheme. Therefore, this invention, based on semantic segmentation image analysis, accurately predicts the position of the crescent bone in the anterior segment of the soft midsection, solving the problems caused by manual annotation, effectively improving the segmentation accuracy and efficiency of the anterior segment of the soft midsection, and is suitable for the needs of modern meat processing production lines.

[0008] Furthermore, the semantic segmentation model is obtained through the following steps: Obtain a batch of soft anterior segment separation lines marked by professionals, which can be used to predict the position of the lunula; An algorithm model is constructed through data training, and new incoming materials generate a soft mid-front separation line based on the algorithm model. Segmentation is performed based on the soft mid-front segment separation line, the actual test results are measured, and the data that exactly cuts to the edge of the lunula is saved as a ground value dataset. Repeat the above process to continuously update the ground truth dataset, and train the semantic segmentation model using all the ground truth datasets.

[0009] Furthermore, both the data and the ground truth dataset were trained using the YOLOv8 object detection model.

[0010] Furthermore, the image acquisition device includes a planar camera, which is used to acquire 2D image information of the original image of the front end of the soft mid-segment to be segmented.

[0011] Furthermore, the semantic segmentation image extracts the segmentation image boundary from the 2D image information and converts it into the segmentation trajectory points.

[0012] Furthermore, the segmentation trajectory points include 6 trajectory points, of which 2 trajectory points are used for soft mid-segment segmentation and 4 trajectory points are used for soft mid-segment segmentation.

[0013] Furthermore, the segmentation trajectory points are converted into coordinate information acceptable to the robotic arm, and the data is sent to the robotic arm. The robotic arm converts the coordinate information to obtain the segmentation trajectory coordinates that the segmentation tool can execute. The robotic arm controls the segmentation tool to segment the soft mid-section and the soft mid-section according to the segmentation trajectory coordinates.

[0014] Furthermore, the segmented trajectory coordinates are obtained through the following steps: The coordinate information is obtained by calculating the angle between two points in the 2D plane to obtain the attitude angle of the segmentation tool movement; The attitude angle and coordinate information are processed and sent to the robotic arm. After the robotic arm obtains the information, it converts the obtained point information according to the set tool coordinate system to obtain the segmented trajectory coordinates.

[0015] A second aspect of the present invention discloses a soft mid-front segmentation system, comprising: A conveyor line, above which a planar camera and a robotic arm are arranged in sequence according to the conveying direction of the conveyor line, and a first photoelectric sensor and a second photoelectric sensor are respectively arranged on the conveyor line below the planar camera and the robotic arm; A planar camera is used to acquire image information from the front end of the software. A robotic arm is used to drive the cutting tools to execute the cutting plan; A host computer is used to store and run the soft front-end segmentation method according to any one of claims 1-8, wherein the host computer is electrically connected to the PLC; The PLC is responsible for triggering detection and receiving results to control physical actions. The PLC is electrically connected to the conveyor line, the robotic arm, the planar camera, the first photoelectric sensor, the second photoelectric sensor, and the dividing tool.

[0016] Furthermore, the cutting tool includes a mounting plate on which a disc blade, a driving device, and a meat-shoveling inclined plate are mounted. The disc blade is connected to the driving device, which drives the disc blade to rotate. The meat-shoveling inclined plate is located below the disc blade and does not contact it, so that when the disc blade cuts the meat, the meat-shoveling inclined plate is used to lift the meat, ensuring that the disc blade cuts the meat completely through.

[0017] Furthermore, the front end of the shovel plate is tapered and has an arc-shaped groove in the middle, and the blade of the disc knife can be embedded in the arc-shaped groove.

[0018] Furthermore, the rear end of the meat-shoveling inclined plate is provided with a pre-separation pressing rod for pressing the dividing line on the meat.

[0019] Furthermore, the mounting plate is respectively provided with a protective cover to prevent meat scraps from splashing and a scraper to clean meat scraps remaining on the disc blade.

[0020] A third aspect of the present invention discloses a soft mid-front segmentation device, comprising: The acquisition module is used to deploy image acquisition devices on the meat processing production line to trigger the acquisition of the original images of the front end of the meat processing module to be segmented; The lunulae prediction module inputs the original image into a pre-trained semantic segmentation model for inference analysis, accurately predicts the position of the lunulae inside the anterior segment of the soft midline, and generates a semantic segmentation image. The segmentation trajectory point determination module plans segmentation trajectory points based on the analysis of semantic segmentation images; The execution module executes the segmentation scheme based on the segmentation trajectory points.

[0021] A fourth aspect of the present invention discloses an electronic device including a processor and a memory, the memory storing computer-executable instructions executable by the processor, the processor executing the computer-executable instructions to implement the above-described soft-front-end segmentation method.

[0022] A fifth aspect of the present invention discloses a computer storage medium storing computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement the aforementioned soft-front-end segmentation method. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart of a soft mid-front segmentation method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the crescent bone according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the segmented trajectory points according to an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the semantic segmentation model formation process according to an embodiment of the present invention; Figure 5 This is a schematic diagram of a soft mid-front segmentation system according to an embodiment of the present invention; Figure 6 This is a schematic diagram of one side of the dividing tool according to an embodiment of the present invention; Figure 7 This is a schematic diagram of the other side of the dividing tool in an embodiment of the present invention; Figure 8 This is a schematic diagram of the soft mid-front segmentation device according to an embodiment of the present invention; Figure 9 This is a schematic diagram of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] like Figure 1 As shown, the first aspect of the present invention discloses a soft mid-front segmentation method, the method comprising: S110, An image acquisition device is deployed on the meat processing production line to acquire the original image of the front section of the soft mid-section to be segmented; wherein, the image acquisition device includes a planar camera, which is used to acquire 2D image information of the original image of the front section of the soft mid-section to be segmented, and the acquired 2D information is used for target semantic segmentation map. S120, the original image is input into a pre-trained semantic segmentation model for inference analysis, accurately predicting the location of the crescent bone within the anterior segment of the soft midline (e.g., ...). Figure 2 The image is generated by analyzing the original image to predict the position of the crescent bone inside the segmented object, thus obtaining the semantic segmentation image. S130, based on the analysis of the semantic segmentation image, plan segmentation trajectory points; wherein, the semantic segmentation image boundaries are extracted from the 2D image information and converted into segmentation trajectory points. In one embodiment, such as Figure 3 As shown, the segmentation trajectory points include 6 trajectory points. Two trajectory points are used for the segmentation of the soft middle and front sections, that is, cutting along the green line. The left side of the green line is the soft middle section, and the right side of the green line is the soft front section. Four trajectory points are used for the segmentation of the soft middle section. That is, after the soft middle section and the soft front section are separated, cutting is performed along the yellow line. The upper part between the yellow line and the green line is No. 3 meat, and the lower part between the yellow line and the green line is pork belly. S140, execute the segmentation plan according to the segmentation trajectory points; first, cut the soft middle section and soft front section according to the green lines marked by the above two trajectory points to complete the separation, and then cut the No. 3 meat and pork belly according to the yellow lines marked by the above four trajectory points to complete the separation.

[0027] Compared with existing technologies, this application has the following advantages: This invention completely replaces manual marking of the crescent bone in the anterior segment of the soft midsection, significantly improving the accuracy and efficiency of crescent bone position marking on meat processing production lines, and adapting to the production needs of modern production lines. By using a semantic segmentation model for identification and marking, the influence of subjective human factors is eliminated, ensuring accurate marking of the crescent bone in the anterior segment of the soft midsection and preventing the risk of accidental damage to the crescent. Simultaneously, it innovatively utilizes a semantic segmentation model to first predict the crescent bone position, and then plans semantic segmentation trajectory points based on semantic segmentation image analysis to form a segmentation scheme. Therefore, this invention, based on semantic segmentation image analysis, accurately predicts the position of the crescent bone in the anterior segment of the soft midsection, solving the problems caused by manual annotation, effectively improving the segmentation accuracy and efficiency of the anterior segment of the soft midsection, and is suitable for the needs of modern meat processing production lines.

[0028] Following the above embodiments, more specifically, as Figure 3 , Figure 4 As shown, the semantic segmentation model is trained through the following steps: Obtain a batch of soft anterior segment separation lines marked by professionals (experienced individuals) to predict the location of the lunula; An algorithm model is built through data training, and new incoming materials generate a soft mid-front separation line based on the algorithm model. Segmentation was performed based on the soft mid-front segment separation line, and the actual test results were measured. Data that exactly cut to the edge of the lunula were saved as the ground value dataset. Repeat the above process, continuously updating the ground truth dataset through manual annotation, segmentation, and comparison, and then train the semantic segmentation model using all the ground truth datasets.

[0029] It should be noted that both the data and the ground truth dataset were trained using the YOLOv8 object detection model.

[0030] Following the above embodiment, more specifically, the segmentation trajectory points are converted into coordinate information acceptable to the robotic arm, and the data is sent to the robotic arm. The robotic arm converts the coordinate information to obtain the segmentation trajectory coordinates that the segmenting tool can execute. The robotic arm controls the segmenting tool to segment the soft mid-section and the soft middle section according to the segmentation trajectory coordinates. The robotic arm is an existing industrial robot that can move in multiple directions with the segmenting tool, thereby enabling the segmenting tool to complete the cutting.

[0031] Furthermore, the segmented trajectory coordinates are obtained through the following steps: The coordinate information is obtained by calculating the angle between two points in the 2D plane to obtain the attitude angle of the segmentation tool movement; The attitude angle and coordinate information are processed and sent to the robotic arm. After receiving the information, the robotic arm converts the acquired point information according to the set tool coordinate system to obtain the segmented trajectory coordinates.

[0032] A second aspect of the present invention discloses a soft mid-front segmentation system, such as Figure 5 As shown, it includes: A planar camera 22 and a robotic arm 23 are arranged sequentially above the conveyor line 21 in the direction of conveying. A first photoelectric sensor 24 and a second photoelectric sensor 25 are respectively arranged on the conveyor line 21 below the planar camera 21 and the robotic arm 23. Planar camera 21 is used to acquire image information from the front end of the software. Robotic arm 23 is used to drive the dividing cutter 26 to execute the dividing scheme; The host computer is used to store and run the above-mentioned software-to-front-end segmentation method. The host computer is electrically connected to the PLC. The PLC is responsible for triggering detection and receiving results to control physical actions. The PLC is electrically connected to the conveyor line 21, the robotic arm 23, the planar camera 21, the first photoelectric sensor 24, the second photoelectric sensor 25, and the dividing cutter 26.

[0033] The workflow of this system is as follows: 1. The conveyor line 21 includes a photo-taking conveyor line 212 and a separation conveyor line 213. The soft mid-section enters below the planar camera 21 under the action of the photo-taking conveyor line 212, triggers the first photoelectric sensor 24 and sends a signal to the PLC. The PLC controls the photo-taking conveyor line 212 to stop, and the PLC sends a photo-taking command to the host computer. The host computer controls the planar camera 21 to complete the image acquisition of the soft mid-section. 2. After the image acquisition is completed, the host computer sends feedback information to the PLC. The PLC controls the image-taking conveyor line 212 to move, so that the raw material moves. When it reaches the position, it triggers the second photoelectric sensor 25 and sends a feedback signal to the PLC. The PLC then controls the image-taking conveyor line 212 to stop moving. 3. Through training the pre-trained soft mid-front-end model, semantic segmentation is performed on the image, the boundary line of the segmented image is extracted, 6 trajectory points on the two-blade segmentation trajectory line are extracted and converted into trajectory point information, the attitude angle of the segmentation tool 26 is calculated by the trajectory algorithm, and the coordinate and attitude information are converted into information acceptable to the robotic arm 23 and sent to the robotic arm 23. The execution segmentation signal is fed back to the PLC. 4. The PLC receives the execution segmentation signal from the host computer, the robotic arm 23 obtains the trajectory signal, executes the trajectory signal, the segmentation cutter 26 segments the front section of the soft middle section, and reaches the waiting position of the second cutter, and sends the signal that the first cut is completed to the PLC; the PLC obtains the signal that the first cut is completed, and controls the separation conveyor line 213 to start, so that the line accelerates to drive the front section to separate. After the second photoelectric sensor 25 detects no sensing signal, it sends the signal that the front section is separated to the robotic arm 23. 5. The robotic arm 23 receives the execution signal and executes the second cutting trajectory. After the second cutting is completed, the pork belly with skin and the No. 3 pork with fat flow into the rear line to complete the unloading. The second photoelectric sensor 25 is triggered, and the PLC receives the cutting completion information and proceeds to cut the next piece of raw material. The next piece of raw material is fed from the front line, and the cycle continues.

[0034] Following the above embodiments, more specifically, as Figure 6 and Figure 7 As shown, the cutting tool 26 includes a mounting plate 1, on which a disc blade 2, a drive unit 3, and a meat-shoveling inclined plate 4 are mounted. The disc blade 2 and the drive unit 3 are connected, and the drive unit 3 drives the disc blade 2 to rotate. The meat-shoveling inclined plate 4 is located below the disc blade 2 and does not contact it. When the disc blade 2 cuts meat, the meat-shoveling inclined plate 4 is used to lift the meat, ensuring that the disc blade 2 completely cuts through the meat. The drive unit 3 includes a servo motor and a reducer. The servo motor drives the disc blade 2 to rotate at high speed through the reducer, thereby achieving the segmentation of soft meat.

[0035] The drive device 3 drives the disc cutter 2 to rotate. When the disc cutter 2 approaches the meat, the meat is first lifted up by the meat scraper 4. At this time, there will be a gap between the meat and the conveyor line. Because of the gap, the disc cutter 2 can not only cut through the meat according to the cutting requirements, but also will not touch the conveyor line.

[0036] Continuing with the above embodiment, more specifically, the front end of the meat-shoveling inclined plate 4 is tapered and has an arc-shaped groove 41 in the middle, into which the blade of the disc knife 2 can be embedded. The tapered front end of the meat-shoveling inclined plate 4 facilitates the shoveling of meat pieces on the conveyor line, making it easier for the meat pieces to be lifted by the meat-shoveling inclined plate 4. Furthermore, the design of the middle groove 41 creates a structure that is high at both ends and low in the middle, preventing the lifted meat pieces from slipping off the meat-shoveling inclined plate 4. Simultaneously, the design of the disc knife 2's blade embedding into the arc-shaped groove 41 creates a gap between the bottom of the meat piece and the meat-shoveling inclined plate 4. This allows the disc knife 2 to cut through the bottom of the meat piece without touching the meat-shoveling inclined plate 4.

[0037] Following the above embodiment, more preferably, the rear end of the meat-shoveling inclined plate 4 is provided with a pre-separation pressing rod 5 for pressing a dividing line on the meat. The separation of the soft front and middle sections of meat includes the separation of the front and middle sections, and the separation of the pork belly and No. 3 meat in the middle section. For example, when separating the pork belly and No. 3 meat in the middle section, the pre-separation pressing rod 5 can be used to press the meat block, clearly separating the pork belly and No. 3 meat (tenderloin) along the texture of the meat. This makes it easier for the robot to recognize the meat block image and thus more accurately determine the cutting trajectory of the disc knife 2.

[0038] Following the above embodiment, more preferably, the mounting plate 1 is provided with a protective cover 7 for preventing meat scraps from splashing and a scraper 8 for cleaning meat scraps remaining on the disc blade 2. The protective cover 7 is positioned above the disc blade 2 to prevent meat scraps or sawdust oil remaining on the disc blade 2 from splashing. The scraper 8 can scrape off the meat scraps remaining on the disc blade 2 to facilitate cleaning of the disc blade 2.

[0039] like Figure 8 As shown, a third aspect of the present invention discloses a soft mid-front segmentation device, comprising: The acquisition module 201 is used to deploy an image acquisition device on the meat processing production line to trigger the acquisition of the original image of the front section of the soft mid-section to be segmented; wherein, the image acquisition device includes a planar camera, which is used to acquire 2D image information of the original image of the front section of the soft mid-section to be segmented, and to acquire 2D information for target semantic segmentation map. The lunulae prediction module 202 inputs the original image into a pre-trained semantic segmentation model for inference analysis, accurately predicts the position of the lunulae inside the soft mid-front segment, and generates a semantic segmentation image; wherein, the semantic segmentation model analyzes the original image to predict the position of the lunulae inside the segmented object, and obtains the semantic segmentation image. The segmentation trajectory point determination module 203 plans segmentation trajectory points based on the analysis of the semantic segmentation image; in one embodiment, such as... Figure 3 As shown, the segmentation trajectory points include 6 trajectory points. Two trajectory points are used for the segmentation of the soft middle and front sections, that is, cutting along the green line. The left side of the green line is the soft middle section, and the right side of the green line is the soft front section. Four trajectory points are used for the segmentation of the soft middle section. That is, after the soft middle section and the soft front section are separated, cutting is performed along the yellow line. The upper part between the yellow line and the green line is No. 3 meat, and the lower part between the yellow line and the green line is pork belly. Execution module 204 executes the segmentation scheme according to the segmentation trajectory points; first, it cuts and separates the soft middle section and soft front section according to the green lines marked by the above two trajectory points, and then cuts and separates the No. 3 meat and pork belly according to the yellow lines marked by the above four trajectory points.

[0040] This device embodiment corresponds to the aforementioned method embodiment and can be understood by referring to each other.

[0041] like Figure 9 As shown, the fourth aspect of the present invention discloses an electronic device. The electronic device 300 provided in the embodiments of this application includes at least: a processor 301, a memory 302, and a computer program stored in the memory 302 and executable on the processor 301. When the processor 301 executes the computer program, it implements the method provided in the embodiments of this application.

[0042] The electronic device 300 provided in this application embodiment may further include a bus 303 connecting different components (including processor 301 and memory 302). The bus 303 represents one or more types of bus structures, including memory bus, peripheral bus, local area bus, etc.

[0043] Memory 302 may include a readable storage medium in the form of volatile memory, such as random access memory (RAM) 3021 and / or cache memory 3022, and may further include read-only memory (ROM) 3023. Memory 302 may also include a program tool 3025 having a set (at least one) of program modules 3024, including but not limited to an operating subsystem, one or more application programs, other program modules, and program data, each or some combination of these examples may include an implementation of a network environment.

[0044] Processor 301 can be a single processing element or a collective term for multiple processing elements. For example, processor 301 can be a central processing unit (CPU) or one or more integrated circuits configured to implement the methods provided in the embodiments of this application. Specifically, processor 301 can be a general-purpose processor, including but not limited to CPUs, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0045] Electronic device 300 can communicate with one or more external devices 304 (e.g., keyboard, remote control, etc.), and also with one or more devices that enable users to interact with electronic device 300 (e.g., mobile phone, computer, etc.), and / or with any device that enables electronic device 300 to communicate with one or more other electronic devices 300 (e.g., router, modem, etc.). This communication can be performed through input / output (I / O) interface 305. Furthermore, electronic device 300 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) through network adapter 306. Figure 9 As shown, network adapter 306 communicates with other modules of electronic device 300 via bus 303. It should be understood that, although... Figure 9 As not shown, other hardware and / or software modules may be used in conjunction with electronic device 300, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, Redundant Arrays of Independent Disks (RAID) subsystems, tape drives, and data backup storage subsystems.

[0046] It should be noted that, Figure 9 The electronic device 300 shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0047] A fifth aspect of the present invention discloses a computer storage medium storing computer-executable instructions. When these computer-executable instructions are invoked and executed by a processor, they cause the processor to implement the aforementioned software-front-end segmentation method. Specifically, the computer-executable instructions can be built into or installed in the processor, so that the processor can implement the method provided in the embodiments of this application by executing the built-in or installed computer-executable instructions.

[0048] Furthermore, the method provided in this application embodiment can also be implemented as a computer program product, which includes program code that implements the method provided in this application embodiment when run on a processor.

[0049] The computer program product provided in this application embodiment may employ one or more computer-readable storage media, which may be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination thereof. Specifically, more specific examples (a non-exhaustive list) of computer-readable storage media include: electrical connections having one or more wires, portable disks, hard disks, RAM, ROM, erasable programmable read-only memory (EPROM), optical fibers, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0050] The computer program product provided in this application embodiment can be a CD-ROM and include program code, and can also run on electronic devices such as computers. However, the computer program product provided in this application embodiment is not limited thereto. In this application embodiment, the computer-readable storage medium can be any tangible medium that contains or stores program code, which can be used by or in conjunction with an instruction execution system, device, or apparatus.

[0051] It should be noted that although several units or sub-units of the device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided and embodied by multiple units.

[0052] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A soft mid-front segmentation method, characterized in that, The method includes: Image acquisition devices are deployed on the meat processing production line to acquire raw images of the front section of the meat to be processed; The original image is input into a pre-trained semantic segmentation model for inference analysis to predict the position of the lunula inside the anterior segment of the soft midline and generate a semantic segmentation image. Based on the analysis of semantic segmentation images, segmentation trajectory points are planned; The segmentation scheme is executed based on the segmentation trajectory points.

2. The soft mid-front segmentation method according to claim 1, characterized in that, The semantic segmentation model is obtained through the following steps: Obtain a batch of soft anterior segment separation lines marked by professionals, which can be used to predict the position of the lunula; An algorithm model is constructed through data training, and new incoming materials generate a soft mid-front separation line based on the algorithm model. Segmentation is performed based on the soft mid-front segment separation line, the actual test results are measured, and the data that exactly cuts to the edge of the lunula is saved as a ground value dataset. Repeat the above process to continuously update the ground truth dataset, and train the semantic segmentation model using all the ground truth datasets.

3. The soft mid-front segmentation method according to claim 2, characterized in that, Both the data and the ground truth dataset were trained using the YOLOv8 object detection model.

4. The soft mid-front segmentation method according to claim 1, characterized in that, The image acquisition device includes a planar camera, which is used to acquire 2D image information of the original image of the front segment of the soft mid-segment to be segmented.

5. The soft mid-front segmentation method according to claim 4, characterized in that, The semantic segmentation image extracts the segmentation image boundary from the 2D image information and converts it into the segmentation trajectory points.

6. The soft mid-front segmentation method according to claim 5, characterized in that, The segmentation trajectory points include 6 trajectory points, of which 2 trajectory points are used for soft mid-segment segmentation and 4 trajectory points are used for soft mid-segment segmentation.

7. A soft mid-front segmentation method according to any one of claims 1-6, characterized in that, The segmentation trajectory points are converted into coordinate information acceptable to the robotic arm, and the data is sent to the robotic arm. The robotic arm converts the coordinate information to obtain the segmentation trajectory coordinates that the segmentation tool can execute. The robotic arm controls the segmentation tool to segment the soft mid-section and the soft mid-section according to the segmentation trajectory coordinates.

8. The soft mid-front segmentation method according to claim 7, characterized in that, The segmented trajectory coordinates are obtained through the following steps: The coordinate information is obtained by calculating the angle between two points in the 2D plane to obtain the attitude angle of the segmentation tool movement; The attitude angle and coordinate information are processed and sent to the robotic arm. After the robotic arm obtains the information, it converts the obtained point information according to the set tool coordinate system to obtain the segmented trajectory coordinates.

9. A soft mid-front-end segmentation system, characterized in that, include: A conveyor line, above which a planar camera and a robotic arm are arranged in sequence according to the conveying direction of the conveyor line, and a first photoelectric sensor and a second photoelectric sensor are respectively arranged on the conveyor line below the planar camera and the robotic arm; A planar camera is used to acquire image information from the front end of the software. A robotic arm is used to drive the cutting tools to execute the cutting plan; A host computer is used to store and run the soft front-end segmentation method according to any one of claims 1-8, wherein the host computer is electrically connected to the PLC; The PLC is responsible for triggering detection and receiving results to control physical actions. The PLC is electrically connected to the conveyor line, the robotic arm, the planar camera, the first photoelectric sensor, the second photoelectric sensor, and the dividing tool.

10. A soft-front-end segmentation system according to claim 9, characterized in that, The cutting tool includes a mounting plate on which a disc blade, a driving device, and a meat-shoveling inclined plate are mounted. The disc blade is connected to the driving device, which drives the disc blade to rotate. The meat-shoveling inclined plate is located below the disc blade and does not contact it, so that when the disc blade cuts the meat, the meat-shoveling inclined plate is used to lift the meat to ensure that the disc blade cuts the meat completely through.

11. A soft mid-front-end segmentation system according to claim 10, characterized in that, The front end of the shovel plate is tapered and has an arc-shaped groove in the middle, and the blade of the disc knife can be embedded in the arc-shaped groove.

12. A soft-front-end segmentation system according to claim 11, characterized in that, The rear end of the meat-shoveling inclined plate is provided with a pre-separation pressing rod for pressing the dividing line on the meat.

13. A soft mid-front segmentation device according to claim 12, characterized in that, The mounting plate is equipped with a protective cover to prevent meat scraps from splashing and a scraper to clean meat scraps remaining on the disc blade.

14. A soft mid-front segmentation device, characterized in that, include: The acquisition module is used to deploy image acquisition devices on the meat processing production line to trigger the acquisition of the original images of the front end of the meat processing module to be segmented; The lunulae prediction module inputs the original image into a pre-trained semantic segmentation model for inference analysis, accurately predicts the position of the lunulae inside the anterior segment of the soft midline, and generates a semantic segmentation image. The segmentation trajectory point determination module plans segmentation trajectory points based on the analysis of semantic segmentation images; The execution module executes the segmentation scheme based on the segmentation trajectory points.

15. An electronic device, characterized in that, It includes a processor and a memory, the memory storing computer-executable instructions that can be executed by the processor, the processor executing the computer-executable instructions to implement the soft front-end segmentation method according to any one of claims 1-8.

16. A computer storage medium, characterized in that, The computer storage medium stores computer-executable instructions, which, when invoked and executed by the processor, cause the processor to implement the soft front-end segmentation method as described in any one of claims 1-8.

Images