A low-illumination image target detection method and device based on color channel transformation enhancement

Abstract

The application discloses a low-illumination image target detection method and device based on color channel transformation enhancement, and belongs to the technical field of image processing and target detection, and comprises the following steps: constructing an image target detection network based on color channel transformation enhancement; introducing learnable parameters to transform different color channel pixel values, and then performing image enhancement; calculating a detection result by using an enhanced image and an original image; calculating a detection loss of the enhanced image and the original image, and judging whether to optimize the image enhancement process by using a regression loss according to the detection loss. The application alleviates the problem of performance decline of a target detection algorithm caused by low image contrast, fuzzy boundary, noise interference and other factors in a low-illumination scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing and target detection, and more particularly to a low-illumination image target detection method and device based on color channel transformation enhancement. BACKGROUND

[0002] Target detection, as an important basic task in the field of computer vision, aims to accurately identify target categories and location information from images, and plays an important role in automatic driving, environmental monitoring, national security and other fields. Due to the wide application demand and the continuous progress of software and hardware technology, deep learning-based target detection technology has attracted widespread attention from academia and industry, and has made great progress. However, the existing learning-based target detection algorithm is mainly optimized on standard datasets such as COCO and PASCAL VOC, and when processing low-illumination images such as night and backlight, the detection performance will be greatly reduced, thereby affecting the judgment ability and decision accuracy of the decision-making system.

[0003] In this context, it is of great theoretical value and practical significance to improve the robustness and generalization of detection algorithms in low-light scenes. Existing low-light enhancement networks can enhance low-light images to make them more consistent with human perception. However, the enhanced results inevitably differ from the actual high-light images, and simply using the enhanced images for perception tasks cannot bring the best gain in visual perception tasks in low-light scenes. To address this problem, some researchers have tried to jointly optimize image enhancement and target detection networks during training. For example, Yin et al. proposed a pyramid enhancement network based on Laplace transform for target detection in low-light environments. Hashmi et al. proposed a feature enhancement method for target detection in low-light conditions, which enhances hierarchical features to improve detection performance. Some research also explores image adaptive solutions. For example, IA-YOLO and GDIP use differentiable image adaptive processing modules to adaptively adjust image contrast, white balance, and perform dehazing, smoothing, sharpening, and other operations to handle target detection tasks in adverse environments. ERUP-YOLO simplifies the classic image processing filter into a pixel-level filter based on Bezier curves and a kernel-based local filter, and does not need to customize the filter combination for specific data. These methods use detection loss to jointly optimize the image adaptive enhancement module and the target detection network, making the processed image more suitable for detection task perception. In addition, MAET and DAINet train the model on low-light synthetic datasets and use multi-task learning and domain adaptation, respectively, to enable the network to extract illumination-invariant information from input images, thereby improving the performance of low-light image target detection. Overall, the current optimization methods for low-light scene target detection have pre-trained image enhancement networks, joint optimization of image enhancement networks, illumination-invariant feature extraction, and other optimization methods, which have achieved significant results. However, they do not fully consider the differences between different color channels. In addition, most existing "enhancement-detection" algorithms use enhanced images for detection tasks, which do not fully utilize the information of the original input images.

[0004] Prior art documents related to the technical background of the present application:

[0005] [1] Yin X, Yu Z, Fei Z, et al. Yin, Xiangchen, et al. PE-YOLO: Pyramidenhancement network for dark object detection[C] / / Proceedings of theInternational Conference on Artificial Neural Networks, 2023: 163-174.

[0006] [2] Hashmi KA, Kallempudi G, Stricker D, Afzal MZ. Featenhancer: Enhancing hierarchical features for object detection and beyond under low-light vision [C] / / Proceedings of the IEEE / CVF International Conference on Computer Vision, 2023: 6725-6735.

[0007] [3] Liu W, Ren G, Yu R, et al. Image-Adaptive YOLO for Object Detection in Adverse Weather Conditions [C] / / Proceedings of the AAAI Conference on Artificial Intelligence, 2022, 36(2): 1792-1800.

[0008] [4] Kalwar S, Patel D, Aanegola A, et al. GDIP: Gated differentiable image processing for object detection in adverse conditions [C] / / Proceedings of IEEE International Conference on Robotics and Automation, 2023: 7083-7089.

[0009] [5] Ogino Y, Shoji Y, Toizumi T et al. ERUP-YOLO: Enhancing Object Detection Robustness for Adverse Weather Condition by Unified Image-Adaptive Processing [EB / OL]. arXiv preprint arXiv:2411.02799, 2024.

[0010] [6]Cui Z, Qi G, Gu L, et al. Multitask aet with orthogonal tangent regularity for dark object detection[C] / / Proceedings of the IEEE / CVF International Conference on Computer Vision, 2021: 2553-2562.

[0011] [7]Du Z, Shi M, Deng J. Boosting object detection with zero-shot day-night domain adaptation[C] / / Proceedings of the IEEE / CVF Conference on Computer Vision and Pattern Recognition, 2024: 12666-12676。 SUMMARY

[0012] The present application aims to overcome the deficiencies of the prior art, and provides a low-illumination image target detection method and device based on color channel transformation enhancement, which improves the detection effect.

[0013] The purpose of the present application is achieved by the following scheme:

[0014] A low-illumination image target detection method based on color channel transformation enhancement, comprising the following steps:

[0015] An image target detection network based on color channel transformation enhancement is constructed, which introduces learnable parameters to transform different color channel pixel values, and then performs image enhancement; the detection result is calculated using the enhanced image and the original image;

[0016] The detection loss of the enhanced image and the original image is calculated, and whether to use the regression loss to optimize the image enhancement process is judged according to the detection loss.

[0017] Further, the image target detection network based on color channel transformation enhancement includes a color channel transformation image enhancement network;

[0018] The color channel transformation image enhancement network includes a color conversion module and an image enhancement network;

[0019] The processing flow of the color conversion module includes: performing color channel transformation on an input image, applying nonlinear operation, combining learned weights with each channel data of the image, and finely adjusting each color channel pixel value; the processing flow of the image enhancement network includes: inputting the output result of the color conversion module into the enhancement network for enhancement, the enhancement network adopts a convolution network in the form of encoding and decoding, and for an input image, multi-scale image features are extracted through a three-layer convolution network, and the multi-scale features are used to obtain an enhanced image through decoding.

[0020] Further, the image target detection network based on color channel transformation enhancement includes a low-illumination image target detection network; a loss function for cooperative optimization of the enhancement network and the target detection network is established , including an original image detection loss, an enhanced image detection loss and a selective regression loss, and is defined as:

[0021] ;

[0022] Among them, is the detection loss of the original input image , is the detection loss of the enhanced image , is the selective self-supervised regression loss, is a balance coefficient.

[0023] Further, the detection loss is defined as:

[0024] ;

[0025] Among them, , and are target category loss, position loss and confidence loss respectively, is a weight parameter of different losses.

[0026] Further, the enhanced image and the original image detection loss are used as the judgment standard, if the enhanced image detection loss is less than the original image detection loss, it indicates that the enhanced model can make the detection effect better; otherwise, it indicates that the detection effect of the enhanced image is not as good as that of the original input image, and further optimization is still needed.

[0027] Further, the selective self-supervised regression loss is introduced when the detection effect of the enhanced image is not as good as that of the original image, L1 loss is introduced, the original image is used as a self-supervised signal, and the image enhancement network is further optimized, and the loss is represented as:

[0028] ;

[0029] Among them, is an enhanced image, and respectively represent the detection loss of the enhanced image and the original input image.

[0030] Further, the original image includes a low-illumination image obtained in a manner of an imaging sensor or image synthesis.

[0031] Further, it further comprises the step of: when obtaining the low-illumination image, labeling the position of the target of interest in the image and annotating the category, for training the low-illumination image target detection network based on the color channel transformation enhancement.

[0032] Further, the low-illumination image target detection network based on the color channel transformation enhancement is based on YOLOv3.

[0033] A low-illumination image target detection device based on color channel transformation enhancement, comprising a processor and a memory, the memory stores a computer program, when the computer program is loaded by the processor, the method as claimed in any one of the above is executed.

[0034] The beneficial effects of the present application include:

[0035] (1) The present application alleviates the problem of performance degradation of target detection algorithm caused by low image contrast, blurred boundary, noise interference and other factors in low-illumination scene, proposes the technical concept of combining enhancement to obtain better image features, and jointly optimizing the enhancement and detection network to utilize the enhancement to empower the detection task, improves the detection effect, and is suitable for low-illumination environment in security monitoring, intelligent driving, night work and other scenes, and overcomes the problem of performance degradation of target detection algorithm caused by low image contrast, blurred boundary, noise interference and other factors in low-illumination scene.

[0036] (2) The present application improves the flexibility of the enhancement strategy by designing a learnable color channel transformation enhancement module to obtain better detection features; the detection loss is used to jointly optimize the enhancement module and the detection network, so that the enhancement network can be optimized in the direction conducive to the detection task; the selective self-supervised regression loss is proposed, and the enhancement network is optimized by self-supervised regression according to the detection result, to further improve the detection effect. BRIEF DESCRIPTION OF DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0038] Figure 1A structure schematic diagram of a low-illumination image target detection network based on color channel transformation enhancement of the present application;

[0039] Figure 2 A structure schematic diagram of a color channel transformation image enhancement network;

[0040] Figure 3 A workflow schematic diagram of a color channel transformation module;

[0041] Figure 4 An effect comparison diagram of the method of the present application and other existing methods. DETAILED DESCRIPTION

[0042] All features disclosed in the embodiments of the present application, or all steps in the methods or processes impliedly disclosed, can be combined and / or extended, replaced, except for mutually exclusive features and / or steps, in any manner.

[0043] The specific implementation process of the present application is as follows:

[0044] The present application aims to solve the following technical problems: existing learning-based target detection algorithms are mainly optimized on standard data sets such as COCO and PASCAL VOC, and when processing low-illumination images such as night and backlight, the detection performance will be greatly reduced, thereby affecting the judgment ability and decision accuracy of the decision system.

[0045] In view of the above technical problems, the present application proposes the following ideas: based on color channel transformation enhancement to realize low-illumination image target detection, through joint optimization of enhancement and detection network, using enhancement processing to promote detection task. By designing a learnable color channel transformation enhancement module, the flexibility of the enhancement strategy is improved, and better detection features are obtained; the detection loss is used to jointly optimize the enhancement module and the detection network, so that the enhancement network can be optimized in the direction conducive to the detection task; a selective self-supervised regression loss is proposed, and the enhancement network is optimized by self-supervised regression according to the detection result, so as to further improve the detection effect.

[0046] Further, in the training stage, the scheme first adjusts the three color channels of the input image through the color conversion module, and then inputs the enhanced image information representation into the detection network; the detection loss of the enhanced image and the original image input into the detection network is calculated to optimize the detection network, and the process optimizes the enhanced network and the detection network together, so that the enhanced network is optimized in the direction more suitable for the perception of the detection network; finally, according to the detection loss of the enhanced image and the original image, it is judged whether to optimize the image enhancement network by using the regression loss. In the inference stage, first, the input image is adaptively enhanced by using the trained enhanced network, so that the image is more suitable for the perception of the detection network; then, the enhanced image is input into the detection network to obtain the detection result containing the target class, position and confidence, and the accurate identification and positioning of the target in the image are realized. The advantage of the scheme is that before the image enhancement module, a color channel conversion step is introduced, the pixel values of the three color channels of the input image are adjusted by skillfully introducing learnable parameters for nonlinear transformation, which greatly improves the adaptability of the model to different lighting conditions and scenes, and makes the model have more flexible adjustment space when processing complex and variable low-light images; at the same time, the selective self-supervised regression loss judges whether to calculate the regression loss for the original image and the enhanced image respectively according to the detection result, which fully ensures that the enhanced image result closely meets the internal needs of the detection task, maximizes the assistance of the enhanced image to the detection network, and improves the low-light image detection effect.

[0047] In more specific implementation details, the present application designs a low-light image target detection network based on color channel transformation enhancement, and the above method is based on the operation of the network. As shown in Figure 1 , the network includes functional networks: (1) color channel transformation image enhancement network; (2) target detection network. The color channel transformation image enhancement network and the target detection network are optimized together, and the color channel transformation image enhancement network is suitable for the fusion and cooperation of mainstream detection networks. The design ideas of specific modules are introduced as follows:

[0048] (1) Color channel transformation image enhancement network

[0049] The problems of low contrast, blurred boundary and noise interference in low-light images affect the detection performance, so a low-light image enhancement network is proposed for image adaptive enhancement, and the structure is as shown in Figure 2 . The low-light image enhancement network of the present application comprises a color conversion module and an image enhancement network composed of a convolutional network. The processing flow of the color conversion module is as shown in Figure 3 . For the input image , the module gives learnable weights to the three color channels respectively These weights are broadcast to the same dimensions as the image space to ensure compatibility with the image data. Then, the pixel values of each channel of the image are weighted through a nonlinear operation. The specific process can be represented as:

[0050] ;

[0051] wherein, represents the pixel value of the input image in the color channel . row column. Through the above processing, the color features of the image can be effectively optimized, and the adaptability of the model to different lighting conditions and scenes can be enhanced.

[0052] After adjusting the image illumination curve by the color conversion module, the image is input into the enhancement network. The network structure is shown in Figure 2 , which adopts a convolutional network in the form of encoding and decoding. The specific parameters of the network are shown in Table 1, wherein C, K, and S represent the number of output channels, the size of the convolution kernel, and the convolution step, respectively. For the input image, first, multi-scale image features are extracted through a three-layer convolutional network, and the specific process is as follows:

[0053] ;

[0054] wherein, represents a convolution module with a kernel of and a step of , represents a Relu activation function. represents the encoded multi-scale features, represents the feature space scale downsampling multiple.

[0055] The extracted multi-scale features are used for decoding to obtain the enhanced image . Specifically, first, the deep layer features are upsampled using deconvolution, and then are concatenated along the channel dimension. Further, convolution and ReLU activation function are used to extract and nonlinearly transform the fused features. Then, the above process is repeated to upsample the concatenated features using deconvolution, and then are concatenated. Finally, the feature information is fused and the channel number is adjusted through a one-layer convolutional network to obtain the enhanced image. The above process can be represented by the formula:

[0056] ;

[0057] ;

[0058] wherein, This indicates feature concatenation along the channel dimension. This represents the output characteristics of the decoding DCR module. Refers to the deconvolution operation process.

[0059] Figure 1 middle and The original image detection results and detection loss are used. and To enhance image detection results and reduce detection loss, This represents the loss from selective self-supervised regression.

[0060] (2) Target Detection Network

[0061] Low-light enhanced images are input into an object detection network for object detection and recognition, obtaining information such as object category and location. The object detection network structure is not strictly limited and can be applied to mainstream end-to-end detection models. Unlike conventional detection models, to optimize the enhancement network in a direction more suitable for the detection network's perception, the original image detection results are used as an aid. A loss function for the collaborative optimization of the enhancement and object detection networks is established, including the original image detection loss, the enhanced image detection loss, and the selective regression loss, defined as:

[0062] ;

[0063] in, Original input image Detection loss, To enhance the image Detection loss, This represents the loss from selective self-supervised regression. This is the balance coefficient.

[0064] Detection loss can be defined as:

[0065] ;

[0066] in, , and These are the target category loss, location loss, and confidence loss in conventional detection. These are the weight parameters for different losses.

[0067] The introduction of the selective self-supervised regression loss is to make the enhanced network output result more suitable for the subsequent detection task. Specifically, the enhanced image detection loss is used as the judgment standard, if the enhanced image detection loss is less than the original image detection loss, it indicates that the enhanced model can make the detection effect better. Otherwise, it indicates that the detection effect of the enhanced image is not as good as the original input image, and further optimization is still needed. When the detection effect of the enhanced image is not as good as the original image, the selective self-supervised regression loss is introduced, the original image is used as a self-supervised signal, and the image enhancement network is further optimized. The loss is represented as:

[0068] ;

[0069] Wherein, is the enhanced image, and represent the detection loss of the enhanced image and the original input image respectively.

[0070] The method of the present application based on the above design network operation steps are as follows: mainly including model training and model reasoning two parts, the following specific use process is introduced:

[0071] S1: a large number of low-illumination images and target labels are obtained by imaging sensor or image synthesis, and are divided into training set and test set;

[0072] S2: a low-illumination image target detection network based on color channel transformation enhancement is built, including a color channel transformation image enhancement network and a target detection network. The innovation point of the network is the color channel transformation image enhancement network, which introduces learnable parameters to transform different color channels, which can improve the flexibility of the enhancement strategy and the adaptability of the enhancement model to different lighting conditions and scenes. The working process of the color channel transformation image enhancement network is:

[0073] 1) the input image is color channel transformed, and the learned weight is combined with the channel data of the image to finely adjust the pixel value of each color channel of the image;

[0074] 2) the output result of the color conversion module is input into the enhancement network for enhancement, first through a three-layer convolutional network to extract multi-scale image features , and then through decoding to obtain the enhanced image .

[0075] S3: Train the low-illumination image target detection network based on color channel transformation enhancement using the training set. The trained model can perform target detection and recognition on input low-illumination images to obtain target category, position, and other information. Further, the trained model is tested using test data. The innovation of the above detection model training lies in the loss function for collaborative optimization of the enhancement network and the target detection network, including the original image detection loss , the enhanced image detection loss , and the selective regression loss . The selective self-supervised regression loss determines whether to calculate the regression loss for the original image and the enhanced image respectively according to the detection result, ensuring that the enhanced image result closely meets the internal requirements of the detection task and maximizes the assistance of the enhanced image to the detection network.

[0076] In other embodiments of the present application, a low-illumination image target detection method based on color channel transformation enhancement is provided based on the above inventive concept, and the specific execution steps are as follows:

[0077] First, obtain a low-illumination image, label the position of the target of interest in the image with a rectangular frame and annotate the category, which is used for training the low-illumination image target detection model based on color channel transformation enhancement.

[0078] Then, build a low-illumination image target detection model based on color channel transformation enhancement. First, prepare the environment required for model building and model training. In this embodiment, the PyTorch coding and running environment is used to implement the low-illumination image target detection model based on color channel transformation enhancement. The model is based on an end-to-end target detection and recognition network structure and mainly consists of a backbone network, a connection neck, a detection head, and the like. In this embodiment, YOLOv3 is used as the basic detection network.

[0079] Then, the three color channels of the input image are adjusted by the color conversion module to input the enhanced image information into the enhancement network. The color conversion module assigns learnable weights to the three color channels and combines the weights with the image channel data using nonlinear operations to weight the image pixel values, effectively optimizing the image color features and enhancing the model's adaptability to different lighting conditions and scenes. In this embodiment, the enhancement network structure and parameters are as shown in Figure 2 and Table 1.

[0080] Further, the enhanced image is input to the detection model for target detection, and the enhanced network and the detection network are jointly optimized by the detection loss, so that the enhancement processing better supports the detection task. In this embodiment, YOLOv3 is used as the basic detection model, and the training loss includes image detection loss, enhanced image detection loss and selective regression loss.

[0081] The labeled low-illumination image is input into the low-illumination image target detection model based on color channel transformation enhancement for training. In the base model training stage, the YOLOv3 network pre-trained on the COCO dataset is used for optimization. The linear learning rate is used, the initial learning rate is 0.01, and the last round is reduced to 0.001. The batch size in the first stage training is set to 8, and the batch size in the second stage training is set to 1. The SGD optimizer is used, the momentum hyperparameter is set to 0.937, the weight decay hyperparameter is set to 0.0005, and the input image size is fixed to 640x640. The trained model is saved and input into the test image for testing to obtain the final target comprehensive detection and recognition result.

[0082] The low-illumination image target detection model based on color channel transformation enhancement effectively uses the enhancement processing to support low-illumination target detection, introduces learnable parameters to transform different color channels, can improve the flexibility of the enhancement strategy, and jointly optimizes the enhancement module and the detection network, so that the enhancement network can be optimized in the direction conducive to the detection task. Based on the selective self-supervised regression loss, the detection result is used to perform self-supervised regression optimization on the enhancement network, and the detection effect is improved. For the problem of detection effect decline caused by low image contrast, blurred boundary and noise interference in low-illumination scene, this model can better solve it, as shown in Figure 4 YOLOv3 is the basic detection algorithm, SCI-YOLOv3 uses the SCI image enhancement network to pre-process the image, thereby improving the image brightness. IAYOLO is a joint optimization of the image adaptive enhancement module and the detection network to improve the night image detection effect. YOLA enhances the robustness of the detection algorithm in the night scene by learning the illumination invariant feature representation.

[0083] Table 1. Enhancement network structure and parameter table

[0084]

[0085] The units described in the embodiments of the present application can be implemented in software or hardware, and the described units can also be arranged in a processor. In some cases, the names of these units do not constitute a limitation on the units themselves.

[0086] According to an aspect of the embodiments of the present application, there is provided a computer program product or computer program, which comprises computer instructions stored in a computer readable storage medium. A processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the computer device performs the method provided in the various optional implementation manners.

[0087] As another aspect, the embodiments of the present application also provide a computer readable medium, which can be included in the electronic device described in the above embodiments, or can exist separately without being assembled into the electronic device. The computer readable medium carries one or more programs, which, when executed by the electronic device, enable the electronic device to implement the method described in the above embodiments.

Claims

1. A low-light image target detection method based on color channel transformation enhancement, characterized in that, The method comprises the following steps: An image target detection network based on color channel transformation enhancement is constructed, which introduces learnable parameters to transform different color channel pixel values, and then performs image enhancement; A detection result is calculated using the enhanced image and the original image; A detection loss of the enhanced image and the original image is calculated, and whether to optimize the image enhancement process using a regression loss is determined according to the detection loss; The image target detection network based on color channel transformation enhancement comprises a color channel transformation image enhancement network. The color channel transformation image enhancement network comprises a color conversion module and an image enhancement network. The processing flow of the color conversion module comprises: performing color channel transformation on an input image, applying nonlinear operation, combining learned weights with image channel data, and finely adjusting pixel values of each color channel of the image; specifically comprising: for the input image , three color channels are respectively given learnable weights , These weights are broadcast to a scale same as the image space dimension, so as to ensure that the weights can be matched with the image channel data; then, the pixel values of each channel of the image are weighted through nonlinear operation, and the specific process is represented as: ； wherein, represents the pixel values of the input image in the color channel in row column Through the processing flow of the color conversion module, the color features of the image are effectively optimized, which can enhance the adaptability of the model to different light conditions and scenes. The processing flow of the image enhancement network comprises: inputting the output result of the color conversion module into the enhancement network for enhancement, and the enhancement network adopts a convolutional network in the form of encoding and decoding. For an input image, multi-scale image features are extracted through three convolutional networks, and the extracted multi-scale features are used to obtain an enhanced image through decoding. The image target detection network based on color channel transformation enhancement comprises a low-illumination image target detection network; a loss function for cooperative optimization of the enhancement network and the target detection network is established , comprising an original image detection loss, an enhanced image detection loss and a selective regression loss, and is defined as: ； in, Original input image Detection loss, To enhance the image Detection loss, For selective self-supervised regression loss, This is the balance coefficient.

2. The low-light image object detection method based on color channel transformation enhancement of claim 1, wherein, The detection loss is defined as: ； wherein, , and are target class loss, position loss and confidence loss, respectively, are weight parameters for different losses.

3. The low-light image object detection method based on color channel transformation enhancement of claim 2, wherein, The detection loss of the enhanced image and the original image is used as a judgment standard. If the detection loss of the enhanced image is less than that of the original image, it indicates that the enhanced model can make the detection effect better. Otherwise, it indicates that the detection effect of the enhanced image is not as good as that of the original input image, and further optimization is still needed.

4. The low-light image object detection method based on color channel transformation enhancement of claim 3, wherein, When the detection effect of the enhanced image is not as good as that of the original image, the selective self-supervised regression loss is introduced, the original image is used as a self-supervised signal to further optimize the image enhancement network, and the loss is represented as: ； wherein, is the enhanced image, and respectively denote the detection loss for the enhanced image and the original input image.

5. The low-light image object detection method based on color channel transformation enhancement of claim 1, wherein, The original image comprises a low-illumination image obtained by an imaging sensor or image synthesis.

6. The low-light image object detection method based on color channel transformation enhancement of claim 5, wherein, The method further comprises the following steps: when obtaining a low-illumination image, the position of a target of interest in the image is labeled and annotated with a category, which is used for training of the low-illumination image target detection network based on color channel transformation enhancement.

7. The low-light image object detection method based on color channel transformation enhancement of claim 6, wherein, The low-illumination image target detection network based on color channel transformation enhancement is realized based on YOLOv3.

8. A low-illumination image target detection device based on color channel transformation enhancement, characterized in that, The method comprises a processor and a memory, and the memory stores a computer program, which, when loaded by the processor, executes the method according to any one of claims 1-7.

Images