Anti-pinch method and device for food processing machinery, electronic equipment and storage medium

Abstract

The invention discloses an anti-pinch method and device for a food processing machine, electronic equipment and a storage medium, and the method comprises the steps: carrying out the image collection of a feed port of the food processing machine, and obtaining the image data of the feed port of the processing equipment; a target detection algorithm and a deep learning model are adopted for the image data to obtain hand similarity and hand coordinates, threshold judgment is carried out twice according to the hand similarity, if the hand similarity is higher than the first-order threshold, it is determined that the hand is a hand, and if the hand similarity is continuously higher than the second-order threshold, it is determined that the hand is a hand; according to a preset hand clamping high-risk area coordinate range, whether the hand coordinates are located in a high-risk area or not is judged, and if the hand coordinates are located in the high-risk area, it is determined that the hand clamping risk exists; and when it is determined that the hand is the hand and the hand clamping risk exists, a shutdown instruction is sent. According to the method, the hand clamping accident is effectively prevented through twice threshold judgment of the hand similarity and in combination with high-risk judgment of hand clamping.

Description

Technical Field

[0001] The invention relates to the field of mechanical control, and in particular to a method, device, electronic equipment and storage medium for preventing pinching of food processing machinery. Background Art

[0002] The mechanical parts of food processing machinery are harmful to operators. The protective devices based on machinery have a certain impact on the production process and are easily removed in actual production. The protective devices based on infrared technology are relatively simple to implement, but the false detection rate of infrared detection is high, especially for hands and processing objects. Therefore, in the process of identifying the risk of hand pinching in food processing machinery, there are technical difficulties in multiple threshold judgments and hand target identification. The primary problem is the contradiction between the accuracy and real-time performance of hand recognition in a single frame image.

[0003] Too high a recognition threshold will lead to missed detection, while too low a threshold will cause false detection. This requires a balance between recognition accuracy and response speed when setting the first-order threshold. Secondly, to avoid misjudgment caused by instantaneous interference, the system adopts a multi-frame continuous judgment mechanism, which brings up the problem of second-order threshold setting. If the second-order threshold is too high, it will prolong the confirmation time and increase safety risks; if it is too low, it may cause false shutdown due to environmental interference, affecting production efficiency.

[0004] The settings of these two thresholds constrain each other and form a chain reaction. At the same time, the spatial position determination of the hand target is also related to the threshold determination, because the accuracy of the hand coordinates directly depends on the reliability of recognition, and the result of the position determination will in turn affect the credibility of the recognition result. Although this multiple determination mechanism improves security, it also increases the complexity of the system. How to optimize the hand recognition speed while ensuring security is a core issue that needs to be addressed. Summary of the invention

[0005] The present invention provides a method, device, electronic equipment and storage medium for preventing pinching of a food processing machine, which solve the problem of low hand recognition speed.

[0006] According to one aspect of the present invention, the present invention provides a method for preventing pinching of food processing machinery, comprising: performing image acquisition on a feed port of the food processing machinery to obtain image data of the feed port of the processing equipment; using a target detection algorithm and a deep learning model to obtain hand similarity and hand coordinates from the image data, performing two threshold judgments based on the hand similarity, the first-order threshold being greater than the second-order threshold, and if the hand similarity is higher than the first-order threshold, it is identified as a hand, and if the hand similarity is continuously higher than the second-order threshold, it is identified as a hand; based on a preset high-risk area coordinate range for hand pinching, judging whether the hand coordinates are located in a high-risk area, and if they are located in the high-risk area, determining that there is a risk of hand pinching; when it is identified as a hand and there is the risk of hand pinching, sending a stop command to the food processing machinery.

[0007] Furthermore, the threshold judgment is performed twice according to the hand similarity, and the first-order threshold is greater than the second-order threshold. If the hand similarity is higher than the first-order threshold, it is determined to be a hand. If the hand similarity is continuously higher than the second-order threshold, it is determined to be a hand. Specifically: if the hand similarity is higher than the first-order threshold, and the hand coordinates are within the danger zone, the first-order valid label and the second-order valid label are set to invalid, the number of the second-order valid labels is set to zero, and it is determined to be a hand; if the hand similarity is higher than the first-order threshold, and the hand coordinates are not within the danger zone, the first-order valid label is set to valid, the number of the second-order valid labels is set to zero; if If the hand similarity is lower than the first-order threshold and higher than the second-order threshold, and the hand coordinates are not within the danger zone, the second-order valid label is set to be valid; if the hand similarity is lower than the first-order threshold and higher than the second-order threshold, and the hand coordinates are within the danger zone, it is determined whether the first-order valid label or the second-order valid label is valid; if both the first-order valid label and the second-order valid label are invalid, the second-order valid label is set to be valid, and the number of the second-order valid labels is increased by one; if the first-order valid label or the second-order valid label is valid, the first-order valid label and the second-order valid label are set to be invalid and are identified as hands.

[0008] Furthermore, when the number of the second-order valid tags is greater than a preset threshold, the first-order threshold is lowered and the number of the second-order valid tags is set to zero.

[0009] Furthermore, before performing two threshold judgments based on the hand similarity, an inspection is performed to check whether the device is working normally, whether the video device for image acquisition exists, and whether the parameters of the first-order threshold and the parameters of the second-order threshold are complete.

[0010] Furthermore, the image acquisition of the feeding port of the food processing machinery to obtain the image data of the feeding port of the processing equipment is specifically as follows: according to preset image acquisition parameters, the image acquisition of the food processing machinery is performed to obtain the equipment image data, and the acquisition parameters include acquisition time interval, image resolution and image area size; the image preprocessing algorithm is used to perform denoising, enhancement and other preprocessing operations on the equipment image data to obtain the preprocessed equipment image data; according to preset image quality assessment rules, the quality of the preprocessed equipment image data is assessed, if the image quality does not meet the requirements, the equipment image data is re-acquired; the qualified quality equipment image data is segmented by an image segmentation algorithm to extract the image data of the feeding port area.

[0011] Furthermore, the image data is subjected to a target detection algorithm and a deep learning model to obtain hand similarity and hand coordinates, specifically: a feature extraction algorithm is used to extract features from the image data to obtain image feature data; the image feature data is input into a pre-trained deep learning model for inference calculation to obtain the hand similarity; and a target detection algorithm is used to calculate the image feature data to obtain the hand coordinates.

[0012] Furthermore, when it is determined to be a hand and there is the risk of pinching the hand, a shutdown command is sent to the food processing machinery. Specifically, when it is determined to be a hand and there is the risk of pinching the hand, shutdown control information is generated, and the shutdown control information is sent to the control output unit. The control output unit converts the shutdown control information into a PLC control signal, and transmits the PLC control signal to the PLC system of the food processing machinery to complete the shutdown control. According to another aspect of the present invention, there is provided an anti-pinch device for food processing machinery, comprising: a data acquisition module, the data acquisition module being used to collect images of a feed port of the food processing machinery, and obtain image data of the feed port of the processing equipment; a hand identification module, the hand identification module being used to use a target detection algorithm and a deep learning model to obtain hand similarity and hand coordinates from the image data, and to perform two threshold judgments based on the hand similarity, wherein the first-order threshold is greater than the second-order threshold, and if the hand similarity is higher than the first-order threshold, it is identified as a hand, and if the hand similarity is continuously higher than the second-order threshold, it is identified as a hand; a risk confirmation module, the risk confirmation module being used to determine whether the hand coordinates are located in a high-risk area according to a preset high-risk area coordinate range for hand pinching, and if they are located in the high-risk area, determining that there is a risk of hand pinching; and an execution module, the execution module being used to send a stop command to the food processing machinery when it is identified as a hand and there is the risk of hand pinching.

[0013] According to another aspect of the present invention, there is provided an electronic device, comprising: at least one processor, and a memory communicatively connected to the at least one processor;

[0014] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute any food processing machine anti-pinch method in the embodiments of the present invention.

[0015] According to another aspect of the present invention, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable a computer to execute any food processing machine anti-pinch method in the embodiments of the present invention.

[0016] According to the technology of the present invention, by collecting images of the feed port, image data is obtained, and the image data is analyzed using a target detection algorithm and a deep learning model to obtain hand similarity and hand coordinates. Based on the two threshold judgments of hand similarity, combined with the preset coordinate range of the high-risk area for hand pinching, it is determined whether there is a risk of hand pinching. When it is determined that it is a hand and there is a risk of hand pinching, a shutdown command is sent to the food processing machinery in a timely manner, which effectively prevents the occurrence of hand pinching accidents and ensures the personal safety of the operator. The present invention improves the safety of food processing machinery and has practical application value.

[0017] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings are used to better understand the present invention and do not constitute a limitation of the present invention.

[0019] Figure 1 is a flow chart of a method for preventing pinching of a food processing machine provided by an embodiment of the present invention;

[0020] Figure 2 It is a flow chart of a specific implementation method of the food processing machinery anti-pinch method provided by an embodiment of the present invention;

[0021] Figure 3 is a schematic structural diagram of an anti-pinch device for a food processing machine provided by an embodiment of the present invention;

[0022] Figure 4 is a schematic diagram of an electronic device and a storage medium according to an embodiment of the present invention;

[0023] Figure 5 It is a component control flow chart of the anti-pinch device for food processing machinery according to an embodiment of the present invention.

[0024] In the figure, 100, anti-pinch device of food processing machinery; 11, data acquisition module; 12, hand identification module; 13, risk confirmation module; 14, execution module; 200, electronic device; 201, computing unit; 202, ROM; 203, RAM; 204, bus; 205, I / O interface; 206, input unit; 207, output unit; 208, storage unit; 209, communication unit. DETAILED DESCRIPTION

[0025] The following is a description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be appreciated by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope of the present invention. Similarly, for clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0026] like Figure 1 As shown, an embodiment of the present invention discloses a method for preventing pinching of food processing machinery, comprising: S1, performing image acquisition on a feed port of a food processing machinery to obtain image data of the feed port of the processing equipment; S2, using a target detection algorithm and a deep learning model to obtain hand similarity and hand coordinates from the image data, performing two threshold judgments based on the hand similarity, wherein the first-order threshold is greater than the second-order threshold, and if the hand similarity is higher than the first-order threshold, it is identified as a hand, and if the hand similarity is continuously higher than the second-order threshold, it is identified as a hand; S3, judging whether the hand coordinates are located in a high-risk area according to a preset coordinate range of a high-risk area for pinching hands, and if they are located in the high-risk area, determining that there is a risk of pinching hands; S4, when it is identified as a hand and there is a risk of pinching hands, sending a stop command to the food processing machinery.

[0027] The food processing machinery anti-pinch method of the present application acquires image data by performing image acquisition on the feed port, and analyzes the image data using a target detection algorithm and a deep learning model to obtain hand similarity and hand coordinates. Based on the two threshold judgments of the hand similarity, combined with the preset coordinate range of the high-risk area for hand pinching, it is determined whether there is a risk of hand pinching. When it is determined that it is a hand and there is a risk of hand pinching, a shutdown command is sent to the food processing machinery in a timely manner, which effectively prevents the occurrence of hand pinching accidents and ensures the personal safety of the operator. The present invention improves the safety of food processing machinery and has practical application value.

[0028] By combining machine vision, deep learning models and target recognition algorithms, image acquisition, target recognition and reasoning logic are combined and optimized, improving the speed of reasoning and the accuracy of target recognition. The present invention retrains the deep learning model, reduces the complexity of the model, improves the specificity of the model, and increases the speed of target recognition.

[0029] In an optional embodiment of the present invention, step S1 specifically comprises: performing image acquisition on the food processing machinery according to preset image acquisition parameters to obtain equipment image data, the acquisition parameters including acquisition time interval, image resolution and image area size; performing denoising, enhancement and other preprocessing operations on the equipment image data using an image preprocessing algorithm to obtain preprocessed equipment image data; performing quality assessment on the preprocessed equipment image data according to preset image quality assessment rules, and if the image quality does not meet the requirements, re-acquiring the equipment image data; segmenting the qualified equipment image data using an image segmentation algorithm to extract image data of the feed port area.

[0030] Specifically, according to the preset image acquisition parameters of 1920x1080 pixels per 5 seconds, the image data of the feed port of the processing equipment is acquired in real time by a high-speed industrial camera. The median filter algorithm is used to denoise the image, and the adaptive histogram equalization algorithm is used to enhance the image contrast to obtain a high-quality image after preprocessing. The image data of the feed port area is accurately segmented by combining the Canny edge detection algorithm and the region growing algorithm. The SIFT feature extraction algorithm is used to extract the scale-invariant features of the feed port area to obtain a 128-dimensional feature vector. The feature vector is input into the pre-trained SSD target detection model for inference to determine whether there is a hand target at an accuracy rate of 97%. Once the hand is detected, the distance between it and the feed port is calculated by the triangulation method. When the distance is less than the safety threshold of 20 cm, a high-level shutdown command is sent to the PLC system through the digital I / O interface.

[0031] like Figure 2As shown, in an optional embodiment of the present invention, step S2 is specifically: using a feature extraction algorithm to extract features from image data to obtain image feature data; inputting the image feature data into a pre-trained deep learning model for inference calculation to obtain hand similarity; using a target detection algorithm to calculate the image feature data to obtain hand coordinates, if the hand similarity is higher than the first-order threshold, and the hand coordinates are within the danger zone, the first-order valid label and the second-order valid label are set to invalid, the number of second-order valid labels is set to zero, and it is identified as a hand; if the hand similarity is higher than the first-order threshold, and the hand coordinates are not within the danger zone, The first-order valid label is set to valid, and the number of second-order valid labels is set to zero; if the hand similarity is lower than the first-order threshold and higher than the second-order threshold, and the hand coordinates are not within the dangerous area, the second-order valid label is set to valid; if the hand similarity is lower than the first-order threshold and higher than the second-order threshold, and the hand coordinates are within the dangerous area, then determine whether the first-order valid label or the second-order valid label is valid; if both the first-order valid label and the second-order valid label are invalid, the second-order valid label is set to valid, and the number of second-order valid labels is increased by one; when the number of second-order valid labels is greater than the preset threshold, the first-order threshold is lowered, and the number of second-order valid labels is set to zero. ; If the first-order valid label or the second-order valid label is valid, the first-order valid label and the second-order valid label are set to invalid and identified as the hand.

[0032] Specifically, the purpose of the feature extraction algorithm is to convert the image data into a more representative and distinguishable feature vector. For example, the HOG (Histogram of Oriented Gradients) algorithm can be used to extract the shape and texture features of the hand. The HOG algorithm describes the features of the image by calculating the histogram of the gradient direction in the local area of ​​the image, and it has a certain robustness to illumination changes and local deformations. Assume that the extracted image feature data is a 128-dimensional vector, which contains various shapes and texture information of the hand. A pre-trained deep learning model can be used to calculate the hand similarity. For example, a convolutional neural network (CNN) pre-trained on a dataset containing a large number of hand images can be used. The extracted 128-dimensional image feature data is input into the CNN model, and the CNN will output a value between 0 and 1, indicating the similarity between the input image and the hand image in the model. Assume that the similarity output by the CNN model is 92, which indicates that the input image is likely to be a hand. The object detection algorithm can be used to calculate the hand coordinates. For example, the YOLO (You Only Look Once) algorithm can be used to detect the position of the hand. The YOLO algorithm can directly predict the bounding box and category probability of the target from the image, and it has a high detection speed. Assume that the YOLO algorithm detects the coordinates of the hand as (x=100, y=200), which indicates that the center of the hand is located at the (100, 200) pixel position of the image. The first-order threshold and the second-order threshold are used to determine the confidence of the hand similarity. Assume that the first-order threshold is 8.5 and the second-order threshold is 6. If the hand similarity is higher than 8.5, it is considered that the possibility of the hand is high. If the hand similarity is between 6 and 8.5, it is considered that the possibility of the hand is low. The danger zone refers to the area in the food processing machinery where hand pinching accidents are prone to occur. Assume that the danger zone is a rectangular area with the coordinates of the upper left corner (x1=50, y1=150) and the coordinates of the lower right corner (x2=150, y2=250).

[0033] Take the following examples as examples: Example 1: The hand similarity is 92, and the hand coordinates are (100, 200). Since 9.2 is greater than the first-order threshold 8.5, and the hand coordinates (100, 200) are in the danger zone, the first-order valid label and the second-order valid label are set to invalid, the number of second-order valid labels is set to zero, and it is determined to be a hand, thereby triggering a shutdown operation. This is done to avoid hand pinching accidents. Example 2: The hand similarity is 9.2, and the hand coordinates are (300, 400). Since 9.2 is greater than the first-order threshold 8.5, but the hand coordinates (300, 400) are not in the danger zone, the first-order valid label is set to valid, and the number of second-order valid labels is set to zero. This means that the hand is detected, but the hand is not in the danger zone, so there is no need to trigger a shutdown operation. Example 3: The hand similarity is 7, and the hand coordinates are (300, 400). Since 7 is between the second-order threshold 6 and the first-order threshold 8.5, and the hand coordinates (300, 400) are not in the danger zone, the second-order valid label is set to valid. This means that the hand may have been detected, but the confidence is low, and the hand is not in the danger zone, so there is no need to trigger a shutdown operation. Example 4: The hand similarity is 7 and the hand coordinates are (100, 200). Since 7 is between the second-order threshold 6 and the first-order threshold 8.5, and the hand coordinates (100, 200) are in the danger zone. Assuming that both the first-order valid label and the second-order valid label were invalid before, the second-order valid label is set to valid, and the number of second-order valid labels is increased by one. This means that the hand may have been detected, with a low confidence, but the hand is in the danger zone, so further observation is required.

[0034] If the number of second-order valid tags continues to increase, it can be judged as a hand and a shutdown operation is triggered. This is done to avoid false triggering of the shutdown while ensuring safety. If the first-order valid tag was valid before, both the first-order valid tag and the second-order valid tag are set to invalid, and it is determined to be a hand, which triggers a shutdown operation. This means that a hand has been detected before, and now a suspected hand is detected again in the danger zone, it is considered to be a hand, and a shutdown operation is triggered to ensure safety. The purpose of setting two-order thresholds and valid tags is to improve the reliability and safety of the system. Through the two-level judgment mechanism, the probability of misjudgment and missed judgment can be effectively reduced, thereby better protecting the safety of operators.

[0035] like Figure 5 As shown, in an optional embodiment of the present invention, step S4 is specifically: when it is determined to be a hand and there is a risk of pinching the hand, shutdown control information is generated, and the shutdown control information is sent to the control output unit, the control output unit converts the shutdown control information into a PLC control signal, and transmits the PLC control signal to the PLC system of the food processing machinery to complete the shutdown control.

[0036] Specifically, the video acquisition unit uses an industrial-grade high-definition camera, such as Hikvision MV-CA060-20GC, with a frame rate set to 60 frames per second and a resolution of 1920x1080 pixels, to collect image information from the feed port of the processing equipment in real time and encode the collected image data into a video stream in H. 264 format. The microprocessor unit uses Nvidia Jetson Nano, which receives the H. 264 video stream from the camera and transmits the video stream to the core processing algorithm through the RTSP protocol. The core processing algorithm runs on a server configured with an Intel Xeon E5-2690v4 processor and 32GB of memory. The core processing algorithm calls a deep learning model based on YOLOv5 to perform target recognition processing on the received video stream. The model has been trained on a dataset containing 10,000 images of food processing scenes with hand annotations, with an accuracy rate of 95%. The recognition result is output in JSON format, for example, {"objects": [{"class": "hand", "confidence": 98,"bbox": [100, 120, 150, 200]}]}. The core processing algorithm simultaneously calls the contour detection algorithm based on OpenCV to extract the hand contour and calculates the coordinates of the center point of the hand, for example (250, 300), as the hand position information. The business processing algorithm receives the target recognition result and the hand position information. If the "hand" category exists in the target recognition result and the confidence is greater than 9, it is judged that there is a hand. The pre-set hand pinching risk area parameter is a rectangular area with the coordinates of the upper left corner (150, 250) and the coordinates of the lower right corner (350, 450). If there are dangerous parts such as conveyor belts and rotating blades in the area, it is judged that there is a hand pinching risk. The business processing algorithm calculates the distance between the center point of the hand (250, 300) and the hand pinching risk area. If the center point of the hand is located in the risk area of ​​hand pinching, that is, 150<=250<=350 and 250<=300<=450, the hand is judged to be in the high risk area of ​​hand pinching. At this time, the business processing algorithm generates shutdown control information, such as "STOP". The control output unit uses Advantech ADAM-4050 to convert the shutdown control information "STOP" into a PLC control signal, for example, setting the digital output channel DO1 to a high level. The PLC control signal is transmitted to the PLC system of the food processing machinery (for example, Siemens S7-1500) through the Modbus TCP protocol. After receiving the shutdown signal, the PLC system immediately stops the operation of the relevant equipment and completes the shutdown control.

[0037] In an optional embodiment of the present invention, before performing two threshold judgments based on hand similarity, an inspection is performed to check whether the device is working normally, whether a video device for image acquisition exists, and whether the parameters of the first-order threshold and the parameters of the second-order threshold are complete.

[0038] like Figure 3 As shown, the food processing machine anti-clamping device 100 may include:

[0039] The data acquisition module 11 is used to collect images of the food processing machine feed port and obtain image data of the food processing machine feed port;

[0040] The hand recognition module 12 is used to use the target detection algorithm and the deep learning model to obtain the hand similarity and hand coordinates of the image data, and perform two threshold judgments according to the hand similarity. The first-order threshold is greater than the second-order threshold. If the hand similarity is higher than the first-order threshold, it is recognized as a hand. If the hand similarity is continuously higher than the second-order threshold, it is recognized as a hand;

[0041] The risk confirmation module 13 is used to determine whether the hand coordinates are located in a high-risk area according to a preset high-risk area coordinate range for hand pinching. If the hand coordinates are located in the high-risk area, it is determined that there is a risk of hand pinching.

[0042] The execution module 14 is used to send a shutdown command to the food processing machine when it is determined that it is a hand and there is a risk of hand pinching.

[0043] The anti-pinch device 100 for food processing machinery retrains the target detection model according to the hand features of the operator during the operation of the noodle press; the development board of the anti-pinch device 100 for food processing machinery adopts Orange Pie ZERO3, and adopts a camera module with 1 million pixels, 720P, and 180 degrees wide angle; the anti-pinch device 100 for food processing machinery deploys the Ubuntu customized operating system, retains the configuration supporting the Orange Pie hardware, and reduces the resource consumption of the system; the anti-pinch device 100 for food processing machinery deploys and optimizes the NCNN lightweight neural network framework, and improves the speed of reasoning by optimizing parameters. The anti-pinch device 100 for food processing machinery deploys and optimizes the NanoDet lightweight target recognition model, and customizes the model to obtain a target detection algorithm suitable for the noodle press. The control output unit of the anti-pinch device 100 for food processing machinery communicates with the PLC system of the noodle press through the UART interface to control the noodle press to stop; the microprocessor unit of the anti-pinch device 100 for food processing machinery compresses the processing results and uploads the compressed information to the cloud platform through WIFI.

[0044] For the description of specific functions and examples of each module and submodule of the device in the embodiment of the present invention, reference can be made to the relevant description of the corresponding steps in the above method embodiment, which will not be repeated here.

[0045] In the technical solution of the present invention, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0046] According to an embodiment of the present invention, the present invention also provides an electronic device, a readable storage medium and a computer program product.

[0047] Figure 4 A schematic block diagram of an example electronic device 200 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0048] like Figure 4 As shown, the device 200 includes a computing unit 201, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 202 or a computer program loaded from a storage unit 208 into a random access memory (RAM) 203. In the RAM 203, various programs and data required for the operation of the device 200 can also be stored. The computing unit 201, the ROM 202, and the RAM 203 are connected to each other via a bus 204. An input / output (I / O) interface 205 is also connected to the bus 204.

[0049] A number of components in the device 200 are connected to the I / O interface 205, including: an input unit 206, such as a keyboard, a mouse, etc.; an output unit 207, such as various types of displays, speakers, etc.; a storage unit 208, such as a disk, an optical disk, etc.; and a communication unit 209, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 209 allows the device 200 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0050] The computing unit 201 can be various general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 201 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 201 performs the various methods and processes described above, such as a food processing machinery anti-clamping method. For example, in some embodiments, a food processing machinery anti-clamping method can be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 208. In some embodiments, part or all of the computer program can be loaded and / or installed on the device 200 via ROM 202 and / or communication unit 209. When the computer program is loaded into RAM 203 and executed by the computing unit 201, one or more steps of a food processing machinery anti-clamping method described above can be performed. Alternatively, in other embodiments, the computing unit 201 may be configured to execute a food processing machine anti-pinch method in any other appropriate manner (eg, by means of firmware).

[0051] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0052] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0053] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of machine-readable storage media would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0054] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0055] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0056] A computer system may include a client and a server. The client and the server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises through computer programs running on respective computers and having a client-server relationship to each other. The server may be a cloud server, a server in a distributed system, or a server combined with a blockchain.

[0057] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution disclosed in the present invention can be achieved, and this document does not limit this.

[0058] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for preventing pinching of food processing machinery, characterized in that: include: Capture images of the food processing machinery feed inlet to obtain image data of the processing equipment feed inlet; The target detection algorithm and the deep learning model are used to obtain the hand similarity and hand coordinates of the image data, and two threshold judgments are performed according to the hand similarity, where the first-order threshold is greater than the second-order threshold. If the hand similarity is higher than the first-order threshold, it is determined to be a hand, and if the hand similarity is continuously higher than the second-order threshold, it is determined to be a hand; According to the preset high-risk hand-pinching area coordinate range, judging whether the hand coordinates are located in the high-risk area, and if located in the high-risk area, determining that there is a hand-pinching risk; When it is identified as a hand and there is the risk of hand pinching, a stop command is sent to the food processing machinery.

2. The method according to claim 1, characterized in that The threshold judgment is performed twice according to the hand similarity, the first-order threshold is greater than the second-order threshold, if the hand similarity is higher than the first-order threshold, it is identified as a hand, and if the hand similarity is continuously higher than the second-order threshold, it is identified as a hand, specifically: If the hand similarity is higher than the first-order threshold and the hand coordinates are within the danger zone, the first-order valid labels and the second-order valid labels are set to invalid, the number of the second-order valid labels is set to zero, and it is identified as a hand; If the hand similarity is higher than the first-order threshold and the hand coordinates are not within the danger zone, the first-order valid label is set as valid, and the number of the second-order valid labels is set to zero; If the hand similarity is lower than the first-order threshold and higher than the second-order threshold, and the hand coordinates are not within the danger zone, the second-order valid label is set as valid; If the hand similarity is lower than the first-order threshold and higher than the second-order threshold, and the hand coordinates are within the danger zone, determining whether the first-order valid label or the second-order valid label is valid; If both the first-order valid tag and the second-order valid tag are invalid, the second-order valid tag is set to be valid, and the number of the second-order valid tags is increased by one; If the first-order valid tag or the second-order valid tag is valid, the first-order valid tag and the second-order valid tag are set to invalid and are identified as a hand.

3. The method according to claim 2, characterized in that When the number of the second-order valid tags is greater than a preset threshold, the first-order threshold is lowered and the number of the second-order valid tags is set to zero.

4. The method according to claim 1, characterized in that: Before performing two threshold judgments based on the hand similarity, an inspection is performed to check whether the device is working normally, whether the video device for image acquisition exists, and whether the parameters of the first-order threshold and the parameters of the second-order threshold are complete.

5. The method according to claim 1, characterized in that The image acquisition of the food processing machine feed inlet to obtain the image data of the processing equipment feed inlet is specifically as follows: Capturing images of the food processing machinery according to preset image acquisition parameters to obtain equipment image data, wherein the acquisition parameters include acquisition time interval, image resolution, and image area size; Using an image preprocessing algorithm to perform preprocessing operations such as denoising and enhancement on the device image data to obtain preprocessed device image data; According to the preset image quality assessment rules, the quality of the pre-processed device image data is assessed. If the image quality does not meet the requirements, the device image data is collected again; The qualified equipment image data is segmented by an image segmentation algorithm, and the image data of the feed port area is extracted.

6. The method according to claim 1, characterized in that The object detection algorithm and the deep learning model are used to obtain the hand similarity and hand coordinates of the image data, specifically: A feature extraction algorithm is used to extract features from the image data to obtain image feature data; the image feature data is input into a pre-trained deep learning model for inference calculation to obtain the hand similarity; and a target detection algorithm is used to calculate the image feature data to obtain the hand coordinates.

7. The method according to claim 1, characterized in that When it is determined to be a hand and there is the risk of pinching the hand, a stop instruction is sent to the food processing machine, specifically: When it is determined to be a hand and there is the risk of pinching the hand, shutdown control information is generated and sent to the control output unit. The control output unit converts the shutdown control information into a PLC control signal and transmits the PLC control signal to the PLC system of the food processing machinery to complete the shutdown control.

8. A food processing machine anti-pinch device, characterized in that: include: A data acquisition module, the data acquisition module is used to collect images of the food processing machine feed port and obtain image data of the processing equipment feed port; A hand recognition module, the hand recognition module is used to use a target detection algorithm and a deep learning model to obtain hand similarity and hand coordinates for the image data, and perform two threshold judgments based on the hand similarity, where the first-order threshold is greater than the second-order threshold. If the hand similarity is higher than the first-order threshold, it is recognized as a hand, and if the hand similarity is continuously higher than the second-order threshold, it is recognized as a hand; A risk confirmation module, the risk confirmation module is used to determine whether the hand coordinates are located in a high-risk area according to a preset high-risk area coordinate range for hand pinching, and if the hand coordinates are located in the high-risk area, it is determined that there is a risk of hand pinching; An execution module is used to send a stop instruction to the food processing machine when it is identified as a hand and there is the risk of hand pinching.

9. An electronic device, characterized in that: include: at least one processor, and a memory communicatively connected to the at least one processor; The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 7.

Images