One-stage target detection method, device and equipment applied to medical detection

Abstract

The embodiment of the invention discloses a first-stage target detection method, device and equipment applied to medical detection, and the method comprises the steps: obtaining a to-be-processed image which is a medical waste bin; inputting the to-be-processed image into a pre-trained target network detection model to obtain a target detection result; wherein the pre-trained target detection model is a target detection model constructed based on YOLOv8, and the pre-trained target detection model comprises a feature extraction layer, a feature fusion layer and a detection head optimization layer; the feature extraction layer is an iRNB-c2f fusion block, the feature fusion layer is a GhostConv convolution layer, and the detection head optimization layer is realized by adopting a dynamic anchor frame and threshold adaptive method.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method, device and equipment for one-stage target detection applied to medical detection. Background Art

[0002] Medical waste management is a key link in the public health system, which can ensure public health and safety, protect the ecological environment, achieve resource optimization and sustainable development, etc.

[0003] In recent years, target detection technology has gradually been applied in the field of medical waste management. In order to cope with the real-time supervision of medical waste in the whole process of medical waste transportation, the target detection model deployed by the camera is used for real-time target detection, which will effectively improve the efficiency of supervision, avoid the loss of medical waste, and count the medical waste boxes. It is mainly used in the temporary storage room of medical waste in the whole process. Through the existing temporary storage room monitoring camera, the packaged boxes in the temporary storage room are detected. In the prior art, the target detection method used for medical waste management has the problems of low detection efficiency and accuracy of the target detection model, too large model, and difficult to deploy in resource-constrained terminal devices.

[0004] Based on this, the embodiments of this specification provide a one-stage target detection method applied to medical detection. Summary of the invention

[0005] The embodiments of this specification provide a target detection method, device and equipment to solve the following technical problems: To solve the above technical problems, the embodiments of this specification are implemented as follows: The embodiment of this specification provides a one-stage target detection method applied to medical detection, and the one-stage target detection method includes: Acquire an image to be processed, where the image to be processed is a medical waste bin; Inputting the image to be processed into a pre-trained target network detection model to obtain a target detection result; in, The pre-trained target detection model is a target detection model built based on YOLOv8, and the pre-trained target detection model includes a feature extraction layer, a feature fusion layer and a detection head optimization layer; the feature extraction layer is an iRMB-c2f fusion block, the feature fusion layer is a GhostConv convolution layer, and the detection head optimization layer is implemented using a dynamic anchor frame and threshold adaptive method.

[0006] The embodiment of this specification provides a first-stage target detection device for medical detection, the target detection device comprising: An acquisition module is used to acquire an image to be processed, wherein the image to be processed is a medical waste bin; A detection module inputs the image to be processed into a pre-trained target network detection model to obtain a target detection result; in, The pre-trained target detection model is a target detection model built based on YOLOv8, and the pre-trained target detection model includes a feature extraction layer, a feature fusion layer and a detection head optimization layer; the feature extraction layer is an iRMB-c2f fusion block, the feature fusion layer is a GhostConv convolution layer, and the detection head optimization layer is implemented using a dynamic anchor frame and threshold adaptive method.

[0007] The embodiment of the present specification also provides an electronic device, including: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: Acquire an image to be processed, where the image to be processed is a medical waste bin; Inputting the image to be processed into a pre-trained target network detection model to obtain a target detection result; in, The pre-trained target detection model is a target detection model built based on YOLOv8, and the pre-trained target detection model includes a feature extraction layer, a feature fusion layer and a detection head optimization layer; the feature extraction layer is an iRMB-c2f fusion block, the feature fusion layer is a GhostConv convolution layer, and the detection head optimization layer is implemented using a dynamic anchor frame and threshold adaptive method.

[0008] At least one of the above-mentioned technical solutions adopted in the embodiments of this specification can achieve the following beneficial effects: the one-stage target detection method applied to medical detection provided in the embodiments of this specification, together with the lightweight model, improves the accuracy and efficiency of target detection, ensures that there will be no false detection or missed detection, and has good robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0010] Figure 1A schematic diagram of the system architecture of a one-stage target detection method applied to medical detection provided in an embodiment of this specification; Figure 2 A flowchart of a first-stage target detection method applied to medical detection provided in an embodiment of this specification; Figure 3 A schematic diagram of the structure of the feature extraction layer provided in the embodiments of this specification; Figure 4 A schematic diagram of a first-stage target detection device for medical detection provided in an embodiment of this specification. DETAILED DESCRIPTION

[0011] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0012] Figure 1 The system architecture diagram of a one-stage target detection method for medical detection provided in the embodiments of this specification is shown in FIG. Figure 1 As shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0013] The terminal devices 101, 102, 103 interact with the server 105 via the network 104 to receive or send messages, etc. Various client applications may be installed on the terminal devices 101, 102, 103. For example, a dedicated application with functions such as first-stage target detection for medical detection, etc.

[0014] The terminal devices 101, 102, 103 may be hardware or software. When the terminal devices 101, 102, 103 are hardware, they may be various dedicated or general electronic devices, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, etc. When the terminal devices 101, 102, 103 are software, they may be installed in the electronic devices listed above. They may be implemented as multiple software or software modules (e.g., multiple software or software modules for providing distributed services), or they may be implemented as a single software or software module.

[0015] The server 105 may be a server that provides various services, such as a backend server that provides services for client applications installed on the terminal devices 101, 102, and 103. For example, the server may be applied to the first-stage target detection of medical detection so as to display the first-stage target detection results on the terminal devices 101, 102, and 103.

[0016] The server 105 may be hardware or software. When the server 105 is hardware, it may be implemented as a distributed server cluster consisting of multiple servers, or as a single server. When the server 105 is software, it may be implemented as multiple software or software modules (e.g., multiple software or software modules for providing distributed services), or as a single software or software module.

[0017] The one-stage target detection method for medical detection provided in the embodiments of this specification can be executed by the server 105, for example, or by the terminal devices 101, 102, 103. Alternatively, the one-stage target detection method for medical detection in the embodiments of the present disclosure can be partially executed by the terminal devices 101, 102, 103, and the other part can be executed by the server 105.

[0018] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0019] Figure 2 The flowchart of a first-stage target detection method for medical detection provided in the embodiment of this specification. From the program perspective, the execution subject of the process can be a program installed on an application server or an application terminal. It can be understood that the method can be executed by any device, equipment, platform, or device cluster with computing and processing capabilities. Figure 2 As shown, the one-stage target detection method includes the following steps: Step S201: Acquire an image to be processed, where the image to be processed is a medical waste bin.

[0020] In the embodiment of this specification, the image to be processed is a packaged box captured by a collection device in a temporary storage room for medical waste. Specifically, the collection device is a surveillance camera, and the packaged box is a medical waste box. Therefore, in the embodiment of this specification, the image to be processed is an image of a medical waste box captured by a surveillance camera in a temporary storage room for medical waste.

[0021] It is particularly noted that the image to be processed in the embodiments of this specification is a broad concept, specifically, it can be a picture directly acquired, or it can be a video. When the image to be processed is a video, each frame of the picture can be used as the image to be processed.

[0022] In the embodiment of the present specification, the image to be processed includes at least one target object, which is a medical waste bin.

[0023] Step S203: inputting the image to be processed into a pre-trained target network detection model to obtain a target detection result; in, The pre-trained target detection model is a target detection model built based on YOLOv8, and the pre-trained target detection model includes a feature extraction layer, a feature fusion layer and a detection head optimization layer; the feature extraction layer is an iRMB-c2f fusion block, the feature fusion layer is a GhostConv convolution layer, and the detection head optimization layer is implemented using a dynamic anchor frame and threshold adaptive method.

[0024] In the embodiments of the present specification, the iRMB-c2f fusion block uses iRMB and c2f for fusion, wherein the iRMB includes a depth-separable convolution branch and a self-attention branch, and the separable convolution branch is fused with the self-attention branch through channel splicing and layer normalization.

[0025] iRMB achieves a unified perspective by reconsidering the effective components of the inverted residual block (IRB) and Transformer, thereby extending the IRB of CNN to an attention-based model. The design goal of iRMB is to achieve efficient use of computing resources and high accuracy while keeping the model lightweight. In this embodiment, the use of iRMB can improve performance in an environment with limited computing power. At the same time, its inverted residual design improves the processing of information flow, allowing long-distance dependencies to be captured and utilized while keeping the model lightweight, which is critical for tasks such as image classification, object detection, and semantic segmentation. Ultimately, the model can run efficiently on resource-constrained devices while maintaining or improving prediction accuracy.

[0026] To further understand iRMB, iRMB is introduced in detail below. iRMB proposes the Meta-Mobile Block, which realizes the modular design of the model through different expansion ratios and efficient operators, making the model more flexible and efficient. Specifically, different expansion ratios and efficient operators are combined to form specific modules, and the meta-mobile module is abstracted from the multi-head self-attention and feedforward network. It realizes modular design through different expansion ratios and efficient operators. This approach allows the model to adjust its capacity as needed without redesigning the entire network. The core idea of ​​the meta-mobile block is to integrate different operations such as convolution, self-attention, etc. into a unified framework in a pluggable manner, thereby improving the efficiency and flexibility of the model. This design allows the model to make a better trade-off between complexity and computational efficiency, especially for applications that need to run under limited resources.

[0027] iRMB is a hybrid network module that combines depthwise separable convolution and self-attention mechanism, where depthwise separable convolution (DW-Conv) is used to capture spatial features, while the attention mechanism is used to capture the global dependencies between features.

[0028] In the embodiment of this specification, there are 5 Conv modules in the feature extraction layer, and the iRMB includes: 3×3 depthwise separable convolution and self-attention mechanism.

[0029] In the embodiments of this specification, iRMB is used to extract local features such as the edge and stacking texture of the medical waste box, thereby improving the accuracy of target detection.

[0030] In this embodiment, the c2f module includes three branches, namely: ordinary convolution for extracting basic features of the image to be processed, which is 1×1; depthwise separable convolution for retaining local details, which is 3×3; and CSP structure, which is used to reduce redundant calculations and improve gradient propagation efficiency.

[0031] In an embodiment of the present specification, the method further comprises: a separable convolution branch and the self-attention branch are residually connected through an identity mapping. It should be noted that the separable convolution branch referred to here refers to a depth-separable convolution in iRMB. After the residual connection, the original information can be retained to avoid gradient disappearance.

[0032] The output of the c2f module and the iRMB output are spliced ​​in the channel dimension to integrate multi-scale and multi-type features. In order to ensure the stability of training, layer normalization is performed and the output of layer normalization is used as the input feature map of the feature fusion layer.

[0033] To further understand the feature extraction layer, Figure 3 This is a schematic diagram of the structure of the feature extraction layer provided in the embodiment of this specification. Figure 3 As shown in Figure 1, the output of the c2f module and the output of the iRMB are used together as the output of the feature extraction layer. The c2f module is used to retain the multi-scale feature extraction capability and extract the features of the image to be detected, and the iRMB is used to solve the problems of missed detection and false detection of the target object in the image to be detected.

[0034] In order to further reduce the number of model parameters without reducing detection accuracy, in the embodiment of this specification, the feature fusion layer uses a GhostConv convolution layer.

[0035] In the embodiment of this specification, the feature fusion layer performs 1×1 convolution on the input feature map to generate a basic feature map, and the input feature map is the output feature map of the feature extraction layer; The basic feature map is linearly transformed to generate a similar feature map; The input feature map is concatenated with the similar feature map to generate an output feature map of a feature fusion layer.

[0036] In the embodiment of this specification, the linear transformation specifically includes: 1×1 convolution, 3×3 depth-wise separable convolution and 5×5 depth-wise separable convolution.

[0037] In the feature fusion layer, the acquisition of similar feature maps can obtain additional feature maps without increasing the amount of calculation and data, which helps to enrich the representation ability of the model and improve the accuracy of the model.

[0038] The number of similar feature maps is based on the number of input channels m and the number of linear transformations s, and the number of similar feature maps n=m×s. In this embodiment, the number of linear transformations is 3, and the linear transformations specifically include: 1×1 convolution, 3×3 depth-separable convolution, and 5×5 depth-separable convolution.

[0039] In order to further improve the accuracy, robustness and real-time performance of the medical waste bin detection in this application, the detection head is further optimized. Specifically, according to the size and dense stacking characteristics of the medical waste bin, the anchor frame parameters are dynamically adjusted through the dynamic anchor frame to reduce redundant candidate frames and improve detection efficiency.

[0040] In the embodiment of the present specification, the dynamic anchor frame adjusts the size and proportion of the anchor frame through k-means clustering and the actual size of the image to be processed.

[0041] Specifically, in this embodiment, k-means cluster analysis is performed according to the size of the medical waste box. During the analysis, k=3.

[0042] In the embodiment of this specification, the threshold adaptive method includes: According to the deployment environment of the acquisition device, the real-time detection threshold is adjusted, and the IoU corresponding to the real-time detection threshold is less than or equal to the preset threshold; The deployment environment of the acquisition device includes: the light intensity, contrast, blurriness and viewing angle deviation of the acquisition device.

[0043] In this embodiment, the preset threshold is 0.5.

[0044] In one embodiment of the present specification, the threshold adaptive method is: The dynamic confidence of the MLP network is used for adjustment. The calculation formula of the dynamic confidence is: T=σ( w 1⋅Light+ w 2⋅Contrast+ b ) Among them, T represents the dynamic confidence, σ is the Sigmoid function, w 1, w 2, b All are learnable parameters. If the confidence of a dynamic anchor box is greater than or equal to the dynamic confidence, the dynamic anchor box is retained; if the confidence of a dynamic anchor box is less than the dynamic confidence, the dynamic anchor box is filtered.

[0045] The one-stage target detection method for medical detection provided in the embodiments of this specification, together with the lightweight model, improves the accuracy and efficiency of target detection, ensures that there will be no false detection or missed detection, and has good robustness.

[0046] The embodiments of this specification provide a one-stage target detection method applied to medical detection. Based on the same idea, the embodiments of this specification also provide a one-stage target detection device applied to medical detection. Figure 4 A schematic diagram of a first-stage target detection device for medical detection provided in an embodiment of this specification, such as Figure 4 As shown, the first-stage target detection device includes: An acquisition module 401 acquires an image to be processed, wherein the image to be processed is a medical waste bin; The detection module 403 inputs the image to be processed into a pre-trained target network detection model to obtain a target detection result; in, The pre-trained target detection model is a target detection model built based on YOLOv8, and the pre-trained target detection model includes a feature extraction layer, a feature fusion layer and a detection head optimization layer; the feature extraction layer is an iRMB-c2f fusion block, the feature fusion layer is a GhostConv convolution layer, and the detection head optimization layer is implemented using a dynamic anchor frame and threshold adaptive method.

[0047] The embodiment of the present specification also provides an electronic device, including: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: Acquire an image to be processed, where the image to be processed is a medical waste bin; Inputting the image to be processed into a pre-trained target network detection model to obtain a target detection result; in, The pre-trained target detection model is a target detection model built based on YOLOv8, and the pre-trained target detection model includes a feature extraction layer, a feature fusion layer and a detection head optimization layer; the feature extraction layer is an iRMB-c2f fusion block, the feature fusion layer is a GhostConv convolution layer, and the detection head optimization layer is implemented using a dynamic anchor frame and threshold adaptive method.

[0048] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0049] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, electronic device, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0050] The apparatus, electronic device, non-volatile computer storage medium and method provided in the embodiments of this specification correspond to each other, and therefore, the apparatus, electronic device and non-volatile computer storage medium also have similar beneficial technical effects as the corresponding method. Since the beneficial technical effects of the method have been described in detail above, the beneficial technical effects of the corresponding apparatus, electronic device and non-volatile computer storage medium will not be repeated here.

[0051] In the 1990s, it was very clear whether the improvement of a technology was hardware improvement (for example, improvement of the circuit structure of diodes, transistors, switches, etc.) or software improvement (improvement of the method flow). However, with the development of technology, many improvements of the method flow today can be regarded as direct improvements of the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved method flow into the hardware circuit. Therefore, it cannot be said that the improvement of a method flow cannot be implemented with hardware entity modules. For example, a programmable logic device (PLD) (such as a field programmable gate array (FPGA)) is such an integrated circuit whose logical function is determined by the user's programming of the device. Designers can "integrate" a digital system on a PLD by programming it themselves, without having to ask chip manufacturers to design and make dedicated integrated circuit chips. Moreover, nowadays, instead of manually making integrated circuit chips, this kind of programming is mostly implemented by "logic compiler" software, which is similar to the software compiler used when developing and writing programs, and the original code before compilation must also be written in a specific programming language, which is called hardware description language (HDL). There is not only one kind of HDL, but many kinds, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used ones are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should also know that it is only necessary to program the method flow slightly in the above-mentioned hardware description languages ​​and program it into the integrated circuit, and then it is easy to obtain the hardware circuit that implements the logic method flow.

[0052] The controller may be implemented in any suitable manner, for example, the controller may take the form of a microprocessor or processor and a computer-readable medium storing a computer-readable program code (e.g., software or firmware) executable by the (micro)processor, a logic gate, a switch, an application-specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller, examples of which include but are not limited to the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320, and the memory controller may also be implemented as part of the control logic of the memory. It is also known to those skilled in the art that, in addition to implementing the controller in a purely computer-readable program code manner, the controller may be implemented in the form of a logic gate, a switch, an application-specific integrated circuit, a programmable logic controller, and an embedded microcontroller by logically programming the method steps. Therefore, such a controller may be considered as a hardware component, and the devices for implementing various functions included therein may also be considered as structures within the hardware component. Or even, the devices for implementing various functions may be considered as both software modules for implementing the method and structures within the hardware component.

[0053] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0054] For the convenience of description, the above devices are described in terms of functions and are divided into various units. Of course, when implementing one or more embodiments of this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0055] Those skilled in the art will appreciate that the embodiments of this specification may be provided as methods, systems, or computer program products. Therefore, the embodiments of this specification may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the embodiments of this specification may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0056] This specification is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data optimization device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data optimization device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0057] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data optimization device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0058] These computer program instructions may also be loaded onto a computer or other programmable data optimization device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0059] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0060] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0061] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0062] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0063] The specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media including storage devices.

[0064] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0065] The above is only an embodiment of this specification and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A one-stage target detection method for medical detection, characterized in that: The one-stage target detection method comprises: Acquire an image to be processed, where the image to be processed is a medical waste bin; Inputting the image to be processed into a pre-trained target network detection model to obtain a target detection result; in, The pre-trained target detection model is a target detection model built based on YOLOv8, and the pre-trained target detection model includes a feature extraction layer, a feature fusion layer and a detection head optimization layer; the feature extraction layer is an iRMB-c2f fusion block, the feature fusion layer is a GhostConv convolution layer, and the detection head optimization layer is implemented using a dynamic anchor frame and threshold adaptive method.

2. The one-stage target detection method according to claim 1, characterized in that: The iRMB-c2f fusion block uses iRMB and c2f for fusion, wherein iRMB includes a depth-separable convolution branch and a self-attention branch, and the separable convolution branch is fused with the self-attention branch through channel splicing and layer normalization.

3. The one-stage target detection method according to claim 2, characterized in that: There are 5 Conv modules in the feature extraction layer, and the iRMB includes: 3×3 depth-separable convolution and self-attention mechanism.

4. The one-stage target detection method according to claim 2, characterized in that: The method further includes: a separable convolution branch and the self-attention branch are residually connected through an identity mapping.

5. The one-stage target detection method according to claim 1, characterized in that: The feature fusion layer performs 1×1 convolution on the input feature map to generate a basic feature map, and the input feature map is the output feature map of the feature extraction layer; The basic feature map is linearly transformed to generate a similar feature map; The input feature map is concatenated with the similar feature map to generate an output feature map of a feature fusion layer.

6. The one-stage target detection method according to claim 5, characterized in that: The linear transformation specifically includes: 1×1 convolution, 3×3 depth-wise separable convolution and 5×5 depth-wise separable convolution.

7. The one-stage target detection method according to claim 1, characterized in that: The dynamic anchor frame adjusts the size and proportion of the anchor frame through k-means clustering and the actual size of the image to be processed.

8. The one-stage target detection method according to claim 1, characterized in that: The threshold adaptive method comprises: According to the deployment environment of the acquisition device, the real-time detection threshold is adjusted, and the IoU corresponding to the real-time detection threshold is less than or equal to the preset threshold; The deployment environment of the acquisition device includes: the light intensity, contrast, blurriness and viewing angle deviation of the acquisition device.

9. A one-stage target detection device for medical detection, characterized in that: The target detection device comprises: An acquisition module is used to acquire an image to be processed, wherein the image to be processed is a medical waste bin; A detection module inputs the image to be processed into a pre-trained target network detection model to obtain a target detection result; in, The pre-trained target detection model is a target detection model built based on YOLOv8, and the pre-trained target detection model includes a feature extraction layer, a feature fusion layer and a detection head optimization layer; the feature extraction layer is an iRMB-c2f fusion block, the feature fusion layer is a GhostConv convolution layer, and the detection head optimization layer is implemented using a dynamic anchor frame and threshold adaptive method.

10. An electronic device comprising: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: Acquire an image to be processed, where the image to be processed is a medical waste bin; Inputting the image to be processed into a pre-trained target network detection model to obtain a target detection result; in, The pre-trained target detection model is a target detection model built based on YOLOv8, and the pre-trained target detection model includes a feature extraction layer, a feature fusion layer and a detection head optimization layer; the feature extraction layer is an iRMB-c2f fusion block, the feature fusion layer is a GhostConv convolution layer, and the detection head optimization layer is implemented using a dynamic anchor frame and threshold adaptive method.

Images