Method and system for detecting small target in dark light environment

Abstract

The invention relates to the technical field of image target detection, and discloses a dark light environment small target detection method and system, and the method comprises the steps: collecting original image information; performing dark light enhancement on the original image; using the improved Zero-DCE network to carry out brightness adjustment on the image after the dark light enhancement; and adding a simAM attention mechanism in a backbone feature extraction module of the yov11, and inputting the enhanced image into the improved YOLOv11 network to obtain a small target detection result. A dark light enhancement training data set is not needed, the data acquisition difficulty is reduced, and the occurrence of dark light enhancement model overfitting in the training process is avoided. The model training speed and the reasoning speed are high, and the requirement for computing power resources of equipment is low. Each pixel of the image has a brightness enhancement curve of the data, and the curves are obtained in training, so that the enhanced image is smoother.

Description

Technical Field

[0001] The present invention relates to the technical field of image target detection, and in particular to a method and system for detecting small targets in a dark environment. Background Art

[0002] At present, traditional image enhancement technologies, such as histogram enhancement or RETINEX-based methods, can improve the brightness and contrast of low-light images to a certain extent, but they often amplify noise at the same time, or lose some key detail information while enhancing the target, and cannot effectively solve the contradiction between low-light and small target detection.

[0003] The target detection algorithm based on deep learning can achieve high accuracy and recall under normal lighting conditions. For example, the PE-Yolo method adds a low-frequency enhancement filter on the basis of the Laplacian pyramid and combines it with yolov3 to achieve end-to-end detection. The low-frequency enhancement filter will introduce noise enhancement while improving the low-frequency detail features. In addition, the backbone network Darknet-53 of yolov3 is prone to gradually lose the detail features of small targets as the number of layers increases during the process of continuous sampling and feature extraction. The performance in the dark light small target detection scenario is still unsatisfactory. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is that the existing dark light enhancement algorithm or dark light detection algorithm is trained based on a large amount of dark light data, and it is difficult to obtain samples.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for detecting small targets in a dark environment, comprising:

[0007] Collecting original image information;

[0008] The improved Zero-DCE network is used to enhance the dark light of the original image;

[0009] Add the simAM attention mechanism to the backbone feature extraction module of yolov11, input the enhanced image into the improved YOLOv11 network, and obtain the small target detection result;

[0010] The improved Zero-DCE network includes, in the basic Zero-DCE network, adding a low-pass enhancement filter to enhance the texture features of the object and filter the noise in the camera imaging.

[0011] As a preferred solution of the method for detecting small targets in a dark light environment of the present invention, the dark light enhancement includes enhancing low light images using Zero-DCE, which is composed of multiple curve estimation modules;

[0012] The curve estimation modules are connected in series, each module is responsible for learning a curve transformation of the input image to enhance the lighting effect of the image;

[0013] For each of the curve estimation modules, it is assumed that the illumination of the image is enhanced by learning a set of pixel-level curves;

[0014] For each pixel in the input low-light image, the goal of the network is to learn a curve so that the pixel value is transformed by the curve to obtain the enhanced pixel value;

[0015] The curve is a brightness enhancement curve, which is automatically learned by the network based on the input low-light image.

[0016] As a preferred solution of the method for detecting small targets in a dark environment of the present invention, the formula of the brightness enhancement curve is:

[0017]

[0018] in, Indicates that the original image I index is ( ), represents the brightness enhancement curve parameters, Indicates The pixel value of the enhanced image; b is an unknown number, representing an equation of any order.

[0019] As a preferred solution of the method for detecting small targets in a dark light environment described in the present invention, the basic Zero-DCE network includes 6 hidden layers and 1 output layer;

[0020] Six of the hidden layers use symmetric skip connections similar to U-Net;

[0021] All layers are ordinary 3x3 equal-length convolutional layers, and the output layer uses tanh as the activation function;

[0022] The improved Zero-DCE network includes: using the output of Zero-DCE layer 3 and the result of layer 3 output after low-frequency processing as the input of layer 4; using the output of layer 4 and the result of layer 2 output after low-frequency processing as the input of layer 5; and using the output of layer 5 and the result of layer 1 output after low-frequency processing as the input of layer 6.

[0023] As a preferred solution of the method for detecting small targets in a dark environment described in the present invention, the filtering operation step includes: setting the improved Zero-DCE intermediate output feature map ,in represents the pixel coordinates of the feature map, i represents the hidden layer index of Zero-DCE;

[0024] For feature maps Perform Fourier transform and convert to frequency domain:

[0025]

[0026] Among them, W represents the width of the feature map, H represents the height of the feature map, represents the coordinates in the frequency domain, Indicates the intensity of the frequency component in the image;

[0027] Use Butterworth low-pass filter BLPF to enhance low-frequency information. The transfer formula of BLPF is:

[0028]

[0029] in, Representing feature points To the center of the characteristic spectrum distance; Represents the cutoff frequency, which is used to control the degree of separation of high and low frequency features. When designing the network structure, an adaptive pooling network is used instead. ; n represents the filter order, which controls the attenuation degree of high-frequency characteristics;

[0030] By multiplying the frequency domain value with the low-pass filter, low-frequency information enhancement is achieved. The low-frequency filter formula is:

[0031]

[0032] in, represents the filtered spectrum;

[0033] Moderately amplify the low-frequency coefficients:

[0034]

[0035] Where k is the enhancement factor, , Represents the amplified filtering result;

[0036] Finally, the inverse Fourier transform obtains the feature map after low-frequency enhancement filtering:

[0037]

[0038] in, Representation feature map After low-frequency filtering, the feature map is obtained.

[0039] As a preferred solution of the method for detecting small targets in a dark light environment described in the present invention, the simAM attention mechanism is set in the feature map, and each pixel position feature is used as a neuron;

[0040] For a given neuron, its output , the output of the surrounding neurons is , which measures the energy of a neuron by calculating the linear combination difference between the neuron and its surrounding neurons :

[0041]

[0042] in, and Indicates the weight coefficient used to balance the difference and correlation degree in energy calculation. and The neuron outputs are and The mean of

[0043] The calculated value of the energy function is converted into an attention weight, and the Sigmod activation function is used as the conversion function. For the output of a given neuron, it is expressed as:

[0044] .

[0045] As a preferred solution of the method for detecting small targets in a dark light environment described in the present invention, the improved YOLOv11 network also includes: after using the dark light enhancement model to enhance the training set image, the enhanced image is input into the improved YOLOv11 network for training, and the network parameters are continuously adjusted so that the network can accurately detect small targets in a dark light.

[0046] A small target detection system in a dark light environment using the method described in the present invention is characterized by:

[0047] A collection unit, collecting original image information;

[0048] The enhancement unit uses the improved Zero-DCE network to perform dark light enhancement on the original image;

[0049] The detection unit adds the simAM attention mechanism to the backbone feature extraction module of yolov11, inputs the enhanced image into the improved YOLOv11 network, and obtains the small target detection result; the improved Zero-DCE network includes, in the basic Zero-DCE network, adding a low-pass enhancement filter to enhance the texture features of the object and filter the noise in the camera imaging.

[0050] A computer device comprises: a memory and a processor; the memory stores a computer program, wherein: the processor implements the steps of any one of the methods of the present invention when executing the computer program.

[0051] A computer-readable storage medium stores a computer program, wherein: when the computer program is executed by a processor, the steps of any one of the methods of the present invention are implemented.

[0052] Beneficial effects of the present invention: The method for detecting small targets in a dark light environment provided by the present invention does not require a dark light enhancement training data set, reduces the difficulty of data acquisition, and avoids overfitting of the dark light enhancement model during the training process. The model training speed and inference speed are fast, and the computing power resources required for the device are low. Each pixel in the image has its own brightness enhancement curve, which is obtained during training, making the enhanced image smoother. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0054] Figure 1 An overall flow chart of a small target detection method in a dark light environment provided by the first embodiment of the present invention;

[0055] Figure 2 A flow chart of a curve enhancement model in a small target detection method in a dark light environment provided by the first embodiment of the present invention;

[0056] Figure 3 A yolov11 detection architecture diagram of a small target detection method in a dark light environment provided by the first embodiment of the present invention;

[0057] Figure 4 A Zero-DCE architecture diagram of a conventional curve enhancement model in a small target detection method in a dark light environment provided by the first embodiment of the present invention;

[0058] Figure 5 A flowchart of a pixel attention mechanism for a small target detection method in a dark light environment provided by the first embodiment of the present invention;

[0059] Figure 6 An original dark-light image of a small target detection method in a dark-light environment provided by the second embodiment of the present invention;

[0060] Figure 7 The pixel distribution of each channel of the original dark light data of a small target detection method in a dark light environment provided by the second embodiment of the present invention;

[0061] Figure 8 An image enhanced by Zero-DCE in a method for detecting a small target in a dark environment provided by a second embodiment of the present invention;

[0062] Fig. 9 The pixel distribution of each channel of the image after the Zero-DCE network enhancement in a small target detection method in a dark light environment provided by the second embodiment of the present invention;

[0063] Fig.10 An image after model enhancement in a method for detecting small targets in a dark environment provided by the second embodiment of the present invention;

[0064] Fig.11 The pixel distribution of each channel of the image after model enhancement in a small target detection method in a dark light environment provided by the second embodiment of the present invention. DETAILED DESCRIPTION

[0065] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0066] Example 1, reference Figure 1-Figure 5 , is an embodiment of the present invention, and provides a method for detecting a small target in a dark environment, comprising:

[0067] S1: Collect original image information.

[0068] S2: Perform dark light enhancement on the original image.

[0069] It should be noted that in a dark environment, the overall brightness of the captured images is generally low, and low-brightness pixels account for a significant proportion. In essence, dark light enhancement aims to increase the pixel value of the foreground of the image, and effectively increase the brightness by increasing the contrast between the foreground and the background, thereby improving the visual effect of the image.

[0070] Because different objects have different reflective properties under different lighting conditions, the pixel values ​​they present in the image show a diverse distribution state. If a linear method is used to enhance the lighting effect of an image, not only will the highly reflective areas be prone to overexposure, but the noise points in the image will also be amplified to the same extent, which will undoubtedly have a serious negative impact on image quality.

[0071] In view of this, a curve model is constructed that can enhance low-brightness pixels to a large extent while maintaining the original state of high-brightness pixels as much as possible, so as to achieve the goal of accurate and effective dark light enhancement. While significantly improving image brightness and contrast, it avoids adverse phenomena such as exposure and noise amplification caused by excessive enhancement, ensuring that the enhanced image can clearly and truly restore the scene information in a dark light environment.

[0072] Zero-DCE (Zero-Reference Deep Curve Estimation) is a deep learning method for low-light image enhancement (such as Figure 4 ), mainly composed of multiple curve estimation modules. It is assumed that the illumination of an image can be enhanced by learning a set of pixel-level curves. For each pixel in the input low-light image, the goal of the network is to learn a suitable curve so that the enhanced pixel value can be obtained after the pixel value is transformed through this curve. These curves are not pre-defined, but are automatically learned by the network based on the input low-light image. These modules are connected in series, and each module is responsible for learning a curve transformation of the input image to gradually enhance the illumination effect of the image.

[0073] Assume that the brightness enhancement curve formula is:

[0074]

[0075] in, Indicates that the original image I index is ( ), represents the brightness enhancement curve parameters, Indicates The pixel value of the enhanced image after order; b is an unknown number, representing an equation of any order. Its purpose is to effectively enhance low-light images without the need for reference images (i.e., without the need for paired low-light and normal-light images for training).

[0076] S3: Use the improved Zero-DCE network to adjust the brightness of the dark light enhanced image, such as Figure 4 .

[0077] Based on the network DCE-Net, a low-pass enhancement filter is added to enhance the texture features of objects and filter the noise in camera imaging. The basic Zero-DCE network is very simple, with a total of 7 layers (6 hidden layers, 1 output layer, and 6 hidden layers use symmetrical skip connections similar to U-Net). All layers are ordinary 3x3 equal-length (stride=1) convolutional layers, and the output layer uses tanh as the activation function, such as Figure 4 shown.

[0078] Since the low-frequency component contains the most image semantic information, in order to enrich the semantic information in the reconstructed image, the present invention adds low-pass filtering processing to the output of the 1st, 2nd and 3rd layers of the Zero-DCE network to ensure the enhancement of low-frequency features. The improved Zero-DCE network includes: the output of the 3rd layer of Zero-DCE and the result of the 3rd layer output after low-frequency processing are used as the input of the 4th layer; the output of the 4th layer and the result of the 2nd layer output after low-frequency processing are used as the input of the 5th layer; the output of the 5th layer and the result of the 1st layer output after low-frequency processing are used as the input of the 6th layer.

[0079] Suppose the improved Zero-DCE intermediate output feature map ,in represents the pixel coordinates of the feature map, i represents the hidden layer index of Zero-DCE; the operation steps are as follows:

[0080] For feature maps Perform Fourier transform and convert to frequency domain:

[0081]

[0082] Among them, W is the width of the feature map, H is the height of the feature map, is the coordinate in the frequency domain, different Corresponding to different frequencies, The amplitude reflects the intensity of the frequency component in the image.

[0083] Use Butterworth low-pass filter BLPF to enhance low-frequency information. The transfer formula of BLPF is:

[0084]

[0085] in, Representing feature points To the center of the characteristic spectrum distance; Represents the cutoff frequency, which is used to control the degree of separation of high and low frequency features. When designing the network structure, an adaptive pooling network is used instead. ; n is the filter order, which controls the degree of attenuation of high-frequency characteristics.

[0086] Finally, the low-frequency information is enhanced by multiplying the frequency domain value with the low-pass filter. The low-frequency filter formula is:

[0087]

[0088] in To convert the original spectrum With the transfer function Multiply them together to get the filtered spectrum. This new spectrum is the result of the Butterworth low-pass filter processing, which achieves the retention and enhancement of low-frequency components (because the low-frequency area ), and at the same time attenuates the high-frequency components (because the high-frequency region and decreases with increasing frequency).

[0089] The low-frequency components are retained and enhanced, while the high-frequency noise is attenuated. In order to further highlight the low-frequency enhancement effect, the low-frequency coefficients are moderately amplified:

[0090]

[0091] Where k is the enhancement factor, , Represents the amplified filtering result.

[0092] Finally, the inverse Fourier transform obtains the feature map after low-frequency enhancement filtering:

[0093]

[0094] in, Representation feature map After low-frequency filtering, the feature map is obtained, where Represents the pixel coordinates of the feature map.

[0095] S4: Add the simAM attention mechanism to the backbone feature extraction module of yolov11, input the enhanced image into the improved YOLOv11 network, and obtain the small target detection result.

[0096] The core of the simAM attention mechanism is to measure the importance of each neuron through an energy function. For a given neuron, its output , the output of the surrounding neurons is (Here the surrounding neurons can be defined according to a certain neighborhood range, such as the area centered on the neuron other neurons in an equirectangular region), measuring its energy by calculating the linear combination difference between it and these surrounding neurons :

[0097]

[0098] in, and Indicates the weight coefficient used to balance the difference and correlation degree in energy calculation. and The neuron outputs are and This linear combination difference calculation method is designed to capture the uniqueness and coordination of neurons in local areas. For example, if a neuron has a large linear combination difference with surrounding neurons, it means that it is more "outstanding" in this local area and may carry some key feature information in the image, such as a mutation representing the edge of an object in the image. At this time, its energy value will be relatively high; on the contrary, if its linear relationship with surrounding neurons is more coordinated and the change trend is similar, its energy value will be lower.

[0099] According to the calculated value of the energy function, it is necessary to further convert it into attention weights so that it can be directly applied to the feature map to adjust the degree of attention to different neurons (that is, different pixel position features). This conversion function plays a role of normalization and mapping, converting the energy value, which is relatively abstract and has an uncertain numerical range, into attention weights so that it falls within a reasonable range (usually between 0 and 1). The present invention uses the Sigmod activation function as the conversion function:

[0100]

[0101] when When the value of is large (i.e., the energy is high), will approach 0, indicating that the attention weight corresponding to the neuron is very low, and will be relatively weakened in subsequent feature weighting operations; When the value of is small (low energy), When it approaches 1, it means that the attention weight of this neuron is high, and its corresponding features will be the focus of attention in subsequent processing, so that the network can adaptively focus on the feature information represented by neurons that are more important in the local area and more consistent with the overall feature consistency, thereby improving the entire network's ability to extract and utilize image features.

[0102] The simAM attention mechanism is added to the backbone feature extraction module of yolov11, and the dark light enhancement model is used to enhance the training set images. The enhanced images are then input into the improved YOLOv11 network for training. By continuously adjusting the network parameters, the network can accurately detect small targets in dark light. Figure 2 , Figure 3 As shown in the figure, the design steps of the curve enhancement model with integrated low-frequency enhancement and the main steps of integrating the attention module based on yolov11 are as follows:

[0103] (1) Create a new switching network at the output of the 2nd and 3rd hidden layers of Zero-DCE.

[0104] (2) Add a channel separation layer to separate the three RGB channels, and perform (3)-(6) processing on each channel:

[0105] (3) Add convolution for filtering and use Gaussian padding.

[0106] (4) Add adaptive pooling and use mean padding.

[0107] (5) Add upsampling.

[0108] (6) Add stitching to stitch the features after RGB processing.

[0109] (7) The output of the second process is concatenated with the output of the fourth layer and given to the sixth layer.

[0110] (8) The processed output of layer 3 is concatenated with the original output and given to layer 5.

[0111] (9) Training curve enhancement model.

[0112] (10) Modify the yolov11backbone structure and add the simAM network after c3k2.

[0113] (11) Modify loss and add NWD loss.

[0114] Among them, a low-frequency filter enhancer is added to the adjacent layer of Zero-DCE to improve the detailed texture features. The yolov11 network architecture with pixel attention mechanism is added. The attention mechanism is added after the C3k2 module of the backbone of yolo to improve the pixel accuracy of segmented objects at different scales.

[0115] On the other hand, this embodiment also provides a small target detection system in a dark light environment, which includes:

[0116] The acquisition unit acquires original image information.

[0117] The enhancement unit,uses the improved Zero-DCE network to perform dark light enhancement on the original image.

[0118] The detection unit adds the simAM attention mechanism to the backbone feature extraction module of yolov11, inputs the enhanced image into the improved YOLOv11 network, and obtains the small target detection result. The improved Zero-DCE network includes adding a low-pass enhancement filter to the basic Zero-DCE network to enhance the texture features of the object and filter the noise in the camera imaging.

[0119] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program code.

[0120] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0121] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0122] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0123] Example 2, reference Figure 6-Figure 11 , which is an embodiment of the present invention, provides a small target detection method in a dark light environment. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0124] Figure 6 : Raw dark light data. Figure 7 : Pixel distribution of raw dark light data. Figure 8 :Zero - DCE enhanced data. Fig. 9 : Pixel distribution of Zero-DCE enhanced data. Fig.10: The model enhanced data provided by the present invention. Fig.11 : Pixel distribution of data after model enhancement provided by the present invention.

[0125] from Figure 7 , Fig. 9 , Fig.11 It can be seen that from the characteristics of the image itself, there is a nonlinear correspondence between the illumination intensity and the pixel value of the image. Specifically, for areas with lower pixel values, since they correspond to lower illumination intensity, the magnification of the area should be increased to increase the brightness of the dark area and enhance the overall visibility of the image; and for areas with higher pixel values, since they already correspond to relatively high illumination intensity, in order to prevent overexposure and overexposure, the magnification of the area needs to be reduced, so as to effectively enhance the overall illumination effect of the dark image and ensure that the brightness of each area of ​​the image is within a reasonable range, avoiding problems such as loss of details and poor visual effects caused by overexposure.

[0126] From the effect after image enhancement, the dark light enhancement method provided by the present invention can effectively improve the illumination of the detailed texture part in the image and eliminate the exposure problem of the area with high illumination intensity in the original data.

[0127] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for detecting small targets in a dark environment, characterized in that: include: Collecting original image information; The improved Zero-DCE network is used to enhance the dark light of the original image; Add the simAM attention mechanism to the backbone feature extraction module of yolov11, input the enhanced image into the improved YOLOv11 network, and obtain the small target detection result; The improved Zero-DCE network includes, in the basic Zero-DCE network, adding a low-pass enhancement filter to enhance the texture features of the object and filter the noise in the camera imaging; The basic Zero-DCE network includes 6 hidden layers and 1 output layer; Six of the hidden layers use symmetric skip connections similar to U-Net; All layers are ordinary 3x3 equal-length convolutional layers, and the output layer uses tanh as the activation function; The improved Zero-DCE network specifically includes: using the output of Zero-DCE layer 3 and the result of layer 3 output after low-frequency processing as the input of layer 4; using the output of layer 4 and the result of layer 2 output after low-frequency processing as the input of layer 5; and using the output of layer 5 and the result of layer 1 output after low-frequency processing as the input of layer 6.

2. The method for detecting small targets in a dark environment as claimed in claim 1, characterized in that: The dark light enhancement includes enhancing low light images using Zero-DCE, which is composed of multiple curve estimation modules; The curve estimation modules are connected in series, each module is responsible for learning a curve transformation of the input image to enhance the lighting effect of the image; For each of the curve estimation modules, it is assumed that the illumination of the image is enhanced by learning a set of pixel-level curves; For each pixel in the input low-light image, the goal of the network is to learn a curve so that the pixel value is transformed by the curve to obtain the enhanced pixel value; The curve is a brightness enhancement curve, which is automatically learned by the network based on the input low-light image.

3. The method for detecting small targets in a dark environment as claimed in claim 2, characterized in that: The brightness enhancement curve formula is: ， in, Indicates that the original image I index is ( ), represents the brightness enhancement curve parameters, Indicates The pixel value of the enhanced image; b is an unknown number, representing an equation of any order.

4. The method for detecting small targets in a dark environment as claimed in claim 3, characterized in that: The steps of the filtering operation include: assuming the improved Zero-DCE intermediate output feature map ,in represents the pixel coordinates of the feature map, i represents the hidden layer index of Zero-DCE; For feature maps Perform Fourier transform and convert to frequency domain: ， Among them, W represents the width of the feature map, H represents the height of the feature map, represents the coordinates in the frequency domain, Indicates the intensity of the frequency component in the image; Use Butterworth low-pass filter BLPF to enhance low-frequency information. The transfer formula of BLPF is: ， in, Representing feature points To the center of the characteristic spectrum distance; Represents the cutoff frequency, which is used to control the degree of separation of high and low frequency features. When designing the network structure, an adaptive pooling network is used instead. ; n represents the filter order, which controls the attenuation degree of high-frequency characteristics; By multiplying the frequency domain value with the low-pass filter, low-frequency information enhancement is achieved. The low-frequency filter formula is: ， in, represents the filtered spectrum; Moderately amplify the low-frequency coefficients: ， Where k is the enhancement factor, , Represents the amplified filtering result; Finally, the inverse Fourier transform obtains the feature map after low-frequency enhancement filtering: ， in, Representation feature map After low-frequency filtering, the feature map is obtained.

5. The method for detecting small targets in a dark environment as claimed in claim 4, characterized in that: The simAM attention mechanism is set in the feature map, and each pixel position feature is used as a neuron; For a given neuron, its output , the output of the surrounding neurons is , which measures the energy of a neuron by calculating the linear combination difference between the neuron and its surrounding neurons : ， in, and Indicates the weight coefficient used to balance the difference and correlation degree in energy calculation. and The neuron outputs are and The mean of The calculated value of the energy function is converted into an attention weight, and the Sigmod activation function is used as the conversion function. For the output of a given neuron, it is expressed as: 。 6. The method for detecting small targets in a dark environment as claimed in claim 5, characterized in that: The improved YOLOv11 network also includes, after using the dark light enhancement model to enhance the training set images, inputting the enhanced images into the improved YOLOv11 network for training, and continuously adjusting the network parameters so that the network can accurately detect small targets in dark light.

7. A small target detection system in a dark light environment using the method according to any one of claims 1 to 6, characterized in that: A collection unit, collecting original image information; The enhancement unit uses the improved Zero-DCE network to perform dark light enhancement on the original image; The detection unit adds the simAM attention mechanism to the backbone feature extraction module of yolov11, inputs the enhanced image into the improved YOLOv11 network, and obtains the small target detection result; The improved Zero-DCE network includes, in the basic Zero-DCE network, adding a low-pass enhancement filter to enhance the texture features of the object and filter the noise in the camera imaging.

8. A computer device comprising: Memory and processor; The memory stores a computer program, wherein the processor implements the steps of any one of the methods of claims 1-6 when executing the computer program.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images