Target detection method and device, electronic equipment and storage medium

Abstract

The application relates to a target detection method and device, electronic equipment and a storage medium, and is applied to the technical field of image processing. The method comprises the following steps: performing feature extraction on a to-be-detected image by using a Backbone part to obtain first image features; processing the first image features by using a Neck part to obtain second image features; the Neck part is obtained by sequentially connecting two convolution layers, one Zoom_cat module and a c2f_ODConv module, two convolution layers, one Zoom_cat module and a c2f_ODConv module, one convolution layer and one connection layer, one c2f_ODConv module, one convolution layer and one connection layer, one c2f_ODConv module, one ScalSeq module and one Add module; and processing the second image features by using a Head part to obtain a target detection result. The accuracy of detection is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, and in particular to a target detection method and device, electronic equipment and a storage medium. BACKGROUND

[0002] In recent years, with the continuous advancement of China's urbanization road, the domestic motor vehicle ownership is increasing, leading to increasing road traffic congestion and accident problems. Fog weather is a common and severe weather environment. In heavy fog weather, the recognition ability of vehicles and pedestrians is greatly reduced due to the influence of heavy fog weather, and the probability of traffic accidents will also increase. Influenced by smoke, smoke, mist and other suspended particulate matters, there are problems such as color distortion, blurring, low contrast, etc. in the imaging process under such environment, making the subsequent classification, tracking, pedestrian re-identification, object detection and other tasks very difficult. Therefore, improving the accuracy of road target detection under foggy weather is of great significance to the development of unmanned driving technology.

[0003] At present, the target detection algorithm based on deep learning model includes two-stage target detection algorithm and single-stage target detection algorithm. The two-stage target detection algorithm mainly divides the target detection into two stages, the first stage generates a series of candidate regions containing the approximate position information of the target, and the second stage classifies and arranges the candidate regions and accurately processes the position. Single-stage target detection algorithm is represented by multi-classification single-stage detector (Single Shot multibox Detector, SSD) and YOLO (You Only Look Once) series. However, the accuracy of the above two methods is relatively low. SUMMARY

[0004] In order to solve the above technical problems, the present application provides a target detection method, device, electronic equipment and storage medium.

[0005] According to a first aspect of the present application, a target detection method is provided, comprising:

[0006] obtaining a to-be-detected image;

[0007] inputting the to-be-detected image into a pre-trained AAO-YOLO detection model; wherein the AAO-YOLO detection model comprises a target Backbone part, a target Neck part and a target Head part;

[0008] performing feature extraction on the to-be-detected image by using the target Backbone part to obtain a first image feature; wherein the target Backbone part comprises an original Backbone part in a YOLOv8 model;

[0009] processing the first image feature by using the target Neck part, to obtain a second image feature; wherein the target Neck part is obtained by sequentially connecting 2 convolutional layers in an original Neck part in a YOLOv8 model, 1 added Zoom_cat module, 1 added c2f_ODConv module, 2 convolutional layers in the original Neck part, 1 added Zoom_cat module, 1 added c2f_ODConv module, 1 convolutional layer in the original Neck part, 1 connection layer in the original Neck part, 1 added c2f_ODConv module, 1 convolutional layer in the original Neck part, 1 connection layer in the original Neck part, 1 added c2f_ODConv module, 1 ScalSeq module in the original Neck part, and 1 Add module in the original Neck part;

[0010] processing the second image feature by using the target Head part, to obtain a target detection result; wherein the target Head part is an original Head part in the YOLOv8 model.

[0011] Optionally, the target Backbone part is obtained by sequentially connecting an AOD-Net module and an original Backbone part in a YOLOv8 model.

[0012] Optionally, the training method of the AAO-YOLO detection model comprises:

[0013] adding 2 Zoom_cat modules and 4 c2f_ODConv modules to an original Neck part in a YOLOv8 model, to construct an initial detection model;

[0014] obtaining a sample image, and determining a target detection result in the sample image;

[0015] training the initial detection model by taking the sample image as input and the target detection result in the sample image as label data, to generate the AAO-YOLO detection model.

[0016] According to a second aspect of the present application, a target detection device is provided, comprising:

[0017] a to-be-detected image acquisition module configured to acquire a to-be-detected image;

[0018] An image to be detected input module is configured to input the image to be detected into a pre-trained AAO-YOLO detection model; wherein the AAO-YOLO detection model comprises a target Backbone part, a target Neck part and a target Head part;

[0019] A first processing module is configured to perform feature extraction on the image to be detected by using the target Backbone part to obtain a first image feature; wherein the target Backbone part comprises an original Backbone part in a YOLOv8 model;

[0020] A second processing module is configured to process the first image feature by using the target Neck part to obtain a second image feature; wherein the target Neck part is obtained by sequentially connecting 2 convolution layers, 1 added Zoom_cat module, 1 added c2f_ODConv module, 2 convolution layers, 1 added Zoom_cat module, 1 added c2f_ODConv module, 1 convolution layer, 1 connection layer, 1 added c2f_ODConv module, 1 convolution layer, 1 connection layer, 1 added c2f_ODConv module, 1 ScalSeq module and 1 Add module in the original Neck part in the YOLOv8 model;

[0021] A third processing module is configured to process the second image feature by using the target Head part to obtain a target detection result; wherein the target Head part is an original Head part in the YOLOv8 model.

[0022] Optionally, the target Backbone part is obtained by sequentially connecting an AOD-Net module and an original Backbone part in a YOLOv8 model.

[0023] Optionally, the target detection device further comprises:

[0024] A model architecture construction module is configured to add 2 Zoom_cat modules and 4 c2f_ODConv modules to an original Neck part in a YOLOv8 model to construct an initial detection model;

[0025] A sample data acquisition module is configured to acquire a sample image and determine a target detection result in the sample image;

[0026] A model training module is configured to train the initial detection model by taking the sample image as input and the target detection result in the sample image as label data, and generate the AAO-YOLO detection model.

[0027] According to a third aspect of the present application, an electronic device is provided, comprising: a processor configured to execute a computer program stored in a memory, the computer program being executed by the processor to implement the method of the first aspect.

[0028] According to a fourth aspect of the present application, a computer readable storage medium is provided, having a computer program stored thereon, the computer program being executed by a processor to implement the method of the first aspect.

[0029] According to a fifth aspect of the present application, a computer program product is provided, which, when running on a computer, causes the computer to execute the method of the first aspect.

[0030] The technical scheme provided by the embodiments of the present application has the following advantages compared with the prior art:

[0031] The YOLO series adopts a structure called a spatial pyramid structure, which divides feature maps into pyramids of different scales. On the basis of the YOLOv8 model, two Zoom_cat modules and four c2f_ODConv modules are added. Among them, the Zoom_cat module, which fuses spatial and scale features, is introduced in the Neck part to enhance the multi-scale information extraction capability of the model. By fusing feature maps of different scales, detailed information is increased. Since small targets may lose details on large-scale feature maps, the Zoom_cat module also introduces a channel and position attention mechanism, focusing on small objects related to information channels and spatial positions to improve detection and segmentation performance. Moreover, the two Zoom_cat modules are both set after two different convolution layers, which can utilize the high-dimensional features of the previous layer for operation to generate new outputs containing more position information and contextual features, enhancing the model's ability to handle complex scenes. The two c2f_ODConv modules are set after the Zoom_cat modules to facilitate the use of feature maps obtained by the previous convolution layers, so that the c2f_ODConv modules can directly access more contextual information to improve feature expression capability, and further optimize and extract features after fusing multi-scale information, thus being able to handle more objects of different sizes. The c2f_ODConv module processes the feature maps that have been spliced, so the two c2f_ODConv modules are set after the connection layers to splice the feature maps directly in the channel dimension, thereby further extracting and optimizing the features in the same scale. By setting the above model structure, the accuracy of target detection can be improved. BRIEF DESCRIPTION OF DRAWINGS

[0032] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and serve to explain the principles of the present application, together with the description.

[0033] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings required to be used in the embodiments or prior art description will be briefly introduced. Obviously, for those skilled in the art, other drawings can also be obtained based on these drawings without creative labor.

[0034] Figure 1 A flow chart of a target detection method in the embodiments of the present application;

[0035] Figure 2 A network structure schematic diagram of an AAO-YOLO detection model in the embodiments of the present application;

[0036] Figure 3 A structure schematic diagram of a target detection device in the embodiments of the present application;

[0037] Figure 4 A structure schematic diagram of an electronic device in the embodiments of the present application. DETAILED DESCRIPTION

[0038] In order to more clearly illustrate the above-mentioned purposes, features and advantages of the present application, the schemes of the present application will be further described below. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0039] In the following description, many specific details are set forth in order to fully understand the present application, but the present application can also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only some of the embodiments of the present application, not all the embodiments.

[0040] The YOLOv8 model has been widely concerned and applied in the field of target detection because of its fast detection speed and high accuracy. However, in the foggy road condition, the image captured by the visual sensor is relatively blurred due to poor visibility, which leads to lack of accuracy and is difficult to meet the accuracy requirement. In view of the above problem, the embodiments of the present application take YOLOv8 as the basic research model, improve and redesign the network structure thereof, and obtain a new AAO-YOLO detection model.

[0041] A Zoom_cat module that integrates spatial and scale features is added to the Neck part to enhance the model's multi-scale information extraction capability. By fusing feature maps of different scales, more detailed information is added. The Zoom_cat module also introduces channel and position attention mechanisms to focus on small objects related to information channels and spatial positions in order to improve detection and segmentation performance.

[0042] In addition, a convolutional module, c2f_ODConv, has been added to the Neck section. It can utilize a novel multidimensional attention mechanism and kernel parallel strategy to learn complementary attention of convolutional kernels along all four dimensions of the kernel space in any convolutional layer, thereby improving detection accuracy.

[0043] See Figure 1 , Figure 1 This is a flowchart of a target detection method in an embodiment of this application, which may include the following steps:

[0044] Step S102: Obtain the image to be detected.

[0045] The image to be detected can be an image taken in hazy weather, and the image quality of the image to be detected is poor.

[0046] Step S104: Input the image to be detected into the pre-trained AAO-YOLO detection model; wherein, the AAO-YOLO detection model includes: target Backbone part, target Neck part and target Head part.

[0047] The AAO-YOLO detection model is built upon the YOLOv8 model. The YOLOv8 model consists of a Backbone, a Neck, and a Head. The Backbone is the foundation of the model, extracting features from the input image; these extracted features form the basis for subsequent network layers in object detection. The Neck, located between the Backbone and Head, is used for feature fusion and enhancement. The Head is the decision-making part of the YOLOv8 model, generating the final detection result. As mentioned earlier, the Zoom_cat and c2f_ODConv modules can be added to the Neck to obtain the target Neck, thereby improving detection accuracy.

[0048] Step S106: Using the target Backbone part, feature extraction is performed on the image to be detected to obtain the first image features; wherein, the target Backbone part includes the original Backbone part in the YOLOv8 model.

[0049] The target Backbone part can be the same as the original Backbone part in the YOLOv8 model, or can be based on the original Backbone part in the YOLOv8 model and increase a dehazing module to improve the target detection performance of the blurred image.

[0050] Optionally, the target Backbone part is obtained by sequentially connecting the AOD-Net module and the original Backbone part in the YOLOv8 model. That is, the dehazing module, that is, the AOD-Net module, is added to the Backbone part. The AOD-Net module uses a multi-scale feature extraction strategy to capture different details of the image through different size convolution kernels, and enhances the expression ability of the model through feature splicing. The purpose is to combine the input blurred image and the extracted features of the network to remove haze and improve the target detection performance of the blurred image.

[0051] In step S108, the first image feature is processed by using the target Neck part to obtain a second image feature. The target Neck part is obtained by sequentially connecting 2 convolution layers in the original Neck part in the YOLOv8 model, 1 added Zoom_cat module, 1 added c2f_ODConv module, 2 convolution layers in the original Neck part, 1 added Zoom_cat module, 1 added c2f_ODConv module, 1 convolution layer in the original Neck part, 1 connection layer in the original Neck part, 1 added c2f_ODConv module, 1 convolution layer in the original Neck part, 1 connection layer in the original Neck part, 1 added c2f_ODConv module, 1 ScalSeq module in the original Neck part, and 1 Add module in the original Neck part.

[0052] Referring to Figure 2 , Figure 2 is a network structure diagram of an AAO-YOLO detection model in an embodiment of the present application. The YOLO series adopts a structure called a spatial pyramid structure, which divides the feature map into different scale pyramids. The Zoom_cat module can capture targets of different sizes. Since small targets may lose details on large-scale feature maps, this is particularly important for small target detection. As can be seen, the two added Zoom_cat modules are respectively arranged after the two convolution layers, so that the high-dimensional features of the previous layer can be used for operation to generate new outputs containing more position information and context features, thereby enhancing the ability of the model to process complex scenes.

[0053] Two c2f ODConv modules are respectively arranged after the Zoom cat module, so as to utilize the feature maps obtained by the front convolution layer, enable the c2f ODConv module to directly access more context information, improve the feature expression capability, further optimize and extract features after fusing multi-scale information, and thus can process more objects with different sizes. The c2f ODConv module processes the feature maps that have been spliced, and thus, the two c2f ODConv modules are respectively arranged after the connection layer, so as to splice the feature maps directly in the channel dimension, thereby further extracting and optimizing the features in the same scale.

[0054] After being processed by the c2f ODConv module, multi-scale feature fusion and 3D convolution operation are performed by the ScalSeq module, so as to enhance the detection capability of the model on different scale targets, and two feature maps are added by the Add module, so as to be suitable for various feature fusion scenes.

[0055] In step S110, the second image features are processed by using the target Head part to obtain a target detection result; wherein the target Head part is an original Head part in the YOLOv8 model.

[0056] The target detection method of the embodiment of the present application introduces the Zoom cat module that fuses spatial and scale features in the Neck part of the YOLOv8 model, can enhance the multi-scale information extraction capability of the model, and increases detailed information by fusing feature maps of different scales. Since small targets may lose details on large scale feature maps, the Zoom cat module further introduces a channel and position attention mechanism, focuses on small objects related to information channels and spatial positions, so as to improve the detection and segmentation performance. Moreover, the two Zoom cat modules are both arranged after two different convolution layers, can utilize high-dimensional features of the previous layer to generate new outputs containing more position information and context features, and enhance the capability of the model to process complex scenes. The c2f ODConv module is further introduced in the Neck part, and the two c2f ODConv modules are respectively arranged after the Zoom cat module, so as to utilize the feature maps obtained by the front convolution layer, enable the c2f ODConv module to directly access more context information, improve the feature expression capability, further optimize and extract features after fusing multi-scale information, and thus can process more objects with different sizes. The c2f ODConv module processes the feature maps that have been spliced, and thus, the two c2f ODConv modules are respectively arranged after the connection layer, so as to splice the feature maps directly in the channel dimension, thereby further extracting and optimizing the features in the same scale. It can be seen that, by arranging the above model structure, the accuracy of target detection can be improved.

[0057] Optionally, the training method of the AAO-YOLO detection model comprises:

[0058] Two Zoom_cat modules and four c2f_ODConv modules are added to the original Neck part in the YOLOv8 model to construct an initial detection model. The network structure of the initial detection model is the same as that of the trained AAO-YOLO detection model, and the network parameter values are different.

[0059] A sample image is obtained, and a target detection result in the sample image is determined.

[0060] In the embodiments of the present application, an RTTS image dataset in real foggy weather can be obtained, which is derived from the RESIDE dataset and contains 4322 real foggy pictures, covering different visibility conditions, such as daytime, cloudy, foggy, hazy, and different weather conditions, as well as different road environments such as urban, rural, and highway, involving a wide range of target objects and multiple categories, and different degrees of occlusion and truncation. The dataset sets five target detection categories, namely cars, people, buses, bicycles, and motorcycles, totaling about 41,000 related targets. The present application intends to divide the dataset in a ratio of 6:3:1, and finally the training set contains 2592 pictures, the validation set contains 1297 pictures, and the test set contains 433 pictures.

[0061] The sample image (i.e., sample set) is input, and the target detection result in the sample image is used as label data to train the initial detection model to generate the AAO-YOLO detection model.

[0062] The performance of the model on unknown data is tested through the validation set, and the model parameters are fine-tuned. The performance of the finally trained AAO-YOLO detection model is tested through the test set.

[0063] When the performance of the AAO-YOLO detection model meets the requirements, comparative experiments and analysis are performed on the model and YOLOv3, YOLOv5, YOLOv7, etc. in the same experimental environment. The experimental results of each model are shown in Table 1.

[0064] Table 1

[0065]

[0066] From the experimental results in Table 1, it can be seen that the average precision mAP of the model in this paper has a significant improvement compared with other models. Compared with YOLOv3, YOLOv5, YOLOv7 and other models, it is increased by 7.1%, 5.9%, 3.6% respectively. Moreover, the model in this paper also has certain improvement effect on the three experimental indexes of precision, recall and mAP5-95. Through the above comparison of experimental results, it is verified that the model in this paper can effectively improve the performance indicators to a certain extent, and has an advantage in comprehensive ability over other models.

[0067] The ablation experiment is used to verify the different effects of the improved methods proposed in this application on the detection performance of the original YOLOv8 algorithm. Eight groups of ablation experiments are designed using the RTTS dataset as the main sample, and the influence of different improved methods on the model detection performance is shown in Table 2.

[0068] Table 2

[0069]

[0070] From Table 2, it can be seen that the precision of the model in this paper reaches 78.1%, and the average precision mAP reaches 72.3%. In the ablation experiment, it can be seen that whether there is an AOD-Net module has a significant effect on the experimental precision and has a great influence on the subsequent improvement. After the YOLOv8 is fused with the Zoom_cat module, the precision and recall are increased by 1.9% and 1.4% respectively. After the C2f_ODConv module is fused, the parameters of the original model are reduced from 8.1 GFLOPs (billion floating point operations per second) to 7.9 GFLOPs. Experiment 8 fuses three modules, which improves the precision, recall and average precision of the original model by 2%, 7% and 4.2% respectively. The final experiment proves that the model has high accuracy compared with the original model, and verifies that the AAO-YOLO detection model can effectively improve the detection ability of foggy road scene pictures.

[0071] It should be noted that although the steps of the method in this application are described in a specific order in the accompanying drawings, this does not require or imply that the steps must be performed in this specific order, or that all the steps shown must be performed to achieve the desired results. In addition or alternatively, some steps can be omitted, a plurality of steps can be combined into one step, and / or one step can be divided into a plurality of steps, etc.

[0072] Corresponding to the above method embodiment, the embodiment of the application also provides a target detection device, see Figure 3 The target detection device 300 comprises:

[0073] The to-be-detected image acquisition module 302 is configured to acquire a to-be-detected image.

[0074] The image to be detected input module 304 is configured to input an image to be detected into a pre-trained AAO-YOLO detection model; the AAO-YOLO detection model comprises a target Backbone part, a target Neck part and a target Head part;

[0075] The first processing module 306 is configured to perform feature extraction on the image to be detected by using the target Backbone part to obtain first image features; the target Backbone part comprises an original Backbone part in a YOLOv8 model;

[0076] The second processing module 308 is configured to process the first image features by using the target Neck part to obtain second image features; the target Neck part is obtained by sequentially connecting 2 convolution layers, 1 added Zoom_cat module, 1 added c2f_ODConv module, 2 convolution layers, 1 added Zoom_cat module, 1 added c2f_ODConv module, 1 convolution layer, 1 connection layer, 1 added c2f_ODConv module, 1 convolution layer, 1 connection layer, 1 added c2f_ODConv module, 1 ScalSeq module and 1 Add module in the original Neck part in the YOLOv8 model;

[0077] The third processing module 310 is configured to process the second image features by using the target Head part to obtain a target detection result; the target Head part is an original Head part in the YOLOv8 model.

[0078] Optionally, the target Backbone part is obtained by sequentially connecting an AOD-Net module and an original Backbone part in the YOLOv8 model.

[0079] Optionally, the target detection device 300 further comprises:

[0080] The model architecture construction module is configured to add 2 Zoom_cat modules and 4 c2f_ODConv modules to the original Neck part in the YOLOv8 model to construct an initial detection model;

[0081] The sample data acquisition module is configured to acquire sample images and determine target detection results in the sample images.

[0082] The model training module is configured to train the initial detection model by taking the sample image as input and the target detection result in the sample image as label data, and generate the AAO-YOLO detection model.

[0083] The specific details of each module or unit in the above device have been described in detail in the corresponding method, and therefore will not be repeated here.

[0084] It should be noted that although several modules or units of the device for action execution are mentioned in the above detailed description, such division is not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into embodied by multiple modules or units.

[0085] The embodiments of the present application also provide an electronic device, referring to Figure 4 , Figure 4 is a structural diagram of the electronic device of the embodiments of the present application, comprising a processor 401, a communication interface 402, a memory 403 and a communication bus 404, wherein the processor 401, the communication interface 402 and the memory 403 complete mutual communication through the communication bus 404;

[0086] The memory 403 is used to store computer programs.

[0087] The processor 401 is used to execute the programs stored in the memory 403, and realize the target detection method described above.

[0088] It should be noted that the communication bus 404 mentioned in the above electronic device can be a PCI (Peripheral Component Interconnect, Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture, Extended Industry Standard Architecture) bus, etc. The communication bus 404 can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, Figure 4 only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0089] The communication interface 402 is used for communication between the above electronic device and other devices.

[0090] The memory 403 can include a RAM (Random Access Memory) and can also include a non-volatile memory such as at least one disk memory. Optionally, the memory 403 can also be at least one storage device located away from the aforementioned processor.

[0091] The processor 401 described above can be a general-purpose processor, including a CPU (Central Processing Unit), an NP (Network Processor), and the like; and can also be a DSP (Digital Signal Processing), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component.

[0092] The embodiment of the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the target detection method.

[0093] It should be noted that the computer readable storage medium shown in the present application can be, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection with one or more conductive wires, a portable computer diskette, a hard disk, a random access memory, a read-only memory, 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 above. In the present application, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or component. The program code contained in the computer readable storage medium can be transmitted by any suitable medium, including but not limited to wireless, wire, optical cable, radio frequency, etc., or any suitable combination of the above.

[0094] In the embodiment of the present application, a computer program product is also provided, which, when running on a computer, enables the computer to execute the target detection method.

[0095] It has to be noted that, in the present document, relational terms are intended only to convey a possible relationship between elements or

[0096] The above description is merely that of a specific implementation to enable a person skilled in the art to understand or implement the application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments described herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A target detection method characterized by, The method comprises: obtaining a to-be-detected image; inputting the to-be-detected image into a pre-trained AAO-YOLO detection model; wherein the AAO-YOLO detection model comprises a target Backbone part, a target Neck part and a target Head part; performing feature extraction on the to-be-detected image by using the target Backbone part to obtain a first image feature; wherein the target Backbone part comprises an original Backbone part in a YOLOv8 model; processing the first image feature by using the target Neck part to obtain a second image feature; wherein the target Neck part is obtained by sequentially connecting 2 convolution layers, 1 added Zoom_cat module, 1 added c2f_ODConv module, 2 convolution layers, 1 added Zoom_cat module, 1 added c2f_ODConv module, 1 convolution layer, 1 connection layer, 1 added c2f_ODConv module, 1 convolution layer, 1 connection layer, 1 added c2f_ODConv module, 1 ScalSeq module and 1 Add module in the original Neck part in the YOLOv8 model; processing the second image feature by using the target Head part to obtain a target detection result; wherein the target Head part is an original Head part in the YOLOv8 model.

2. The method of claim 1, wherein, The target Backbone part is obtained by sequentially connecting an AOD-Net module and an original Backbone part in a YOLOv8 model.

3. The method of claim 1, wherein, The training method of the AAO-YOLO detection model comprises: adding 2 Zoom_cat modules and 4 c2f_ODConv modules to an original Neck part in a YOLOv8 model to construct an initial detection model; obtaining a sample image and determining a target detection result in the sample image; training the initial detection model by taking the sample image as input and the target detection result in the sample image as label data to generate the AAO-YOLO detection model.

4. A target detection apparatus characterized by comprising: The device comprises: a to-be-detected image acquisition module configured to acquire a to-be-detected image; a to-be-detected image input module configured to input the to-be-detected image into a pre-trained AAO-YOLO detection model; wherein the AAO-YOLO detection model comprises a target Backbone part, a target Neck part and a target Head part; The first processing module is configured to perform feature extraction on the image to be detected by using the target Backbone part to obtain first image features, wherein the target Backbone part comprises an original Backbone part in a YOLOv8 model. The second processing module is configured to process the first image features by using the target Neck part to obtain second image features, wherein the target Neck part is obtained by sequentially connecting 2 convolution layers, 1 added Zoom_cat module, 1 added c2f_ODConv module, 2 convolution layers, 1 added Zoom_cat module, 1 added c2f_ODConv module, 1 convolution layer, 1 connection layer, 1 added c2f_ODConv module, 1 convolution layer, 1 connection layer, 1 added c2f_ODConv module, 1 ScalSeq module and 1 Add module in the original Neck part in the YOLOv8 model. The third processing module is configured to process the second image features by using the target Head part to obtain a target detection result, wherein the target Head part is an original Head part in the YOLOv8 model.

5. The apparatus of claim 4, wherein, The target Backbone part is obtained by sequentially connecting an AOD-Net module and an original Backbone part in a YOLOv8 model.

6. The apparatus of claim 4, wherein, The device further comprises: A model architecture construction module is configured to add 2 Zoom_cat modules and 4 c2f_ODConv modules to an original Neck part in a YOLOv8 model to construct an initial detection model. A sample data acquisition module is configured to acquire sample images and determine target detection results in the sample images. A model training module is configured to train the initial detection model by taking the sample images as input and the target detection results in the sample images as label data to generate the AAO-YOLO detection model.

7. An electronic device, comprising: The computer program is executed by the processor to implement the method of any one of claims 1-3. The computer program is executed by the processor to implement the method of any one of claims 1-3.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the method of any one of claims 1-3.

Images