Difficulties In Application Research Articles

DEDBNet: DoG-enhanced dual-branch object detection network for remote sensing object detection

With the improvement of spatial resolution of remote sensing images, object detection of remote sensing images has gradually become a difficult task. Extracted object features are usually hidden in a large amount of interference information in the background due to the complexity and large area of backgrounds, as well as the multi-scale nature of objects in remote sensing images. Still, many existing background weakening methods face difficulties in practical applications and are prone to high rates of false positives and false negatives. Therefore, remote sensing object detection has become increasingly challenging. To address these challenges, a novel background weakening method called Difference of Gaussian (DoG) to weaken background (DWB) module is proposed. Then, we develop a dual-branch network, named DoG-Enhanced Dual-Branch Object Detection Network (DEDBNet) for Remote Sensing Object Detection. The base branch network is responsible for detecting objects, while the DWB's branch network corrects the detected objects using feature-level attention. To combine the features of these branches, we propose two new methods Self-Mutual-Correcter with Detect heads (SMCD) for corrective learning and Map Channel Attention (MCA) for channel attention. Self-Corrector (SC) enables modification and integration of features, while the Mutual-Corrector (MC) enhances the features and further fuses them. We evaluate our proposed network, DEDBNet, through extensive experiments on four public datasets (DOTA with an mAP of 0.836, DIOR with an mAP of 0.871, NWPU VHR-10 with an mAP of 0.973, and RSOD with an mAP of 0.975). The results demonstrate that our method outperforms other state-of-the-art object detection methods significantly for remote sensing images.

Enhanced wavelet based spatiotemporal fusion networks using cross-paired remote sensing images

Spatiotemporal fusion can provide remote sensing images with both high temporal and high spatial resolution for earth observation applications. Most of the state-of-the-art models require three or even five images as input, which may lead to difficulties in practical applications due to bad weather or data missing. In this paper, the enhanced cross-paired wavelet based spatiotemporal fusion networks (ECPW-STFN) are proposed to improve the accuracy and quality of the fusions with fewer remote sensing images as inputs. Wavelet transform is introduced into spatiotemporal image fusion to achieve separate training of the high and low frequency components of the image to better extract features of different levels. In addition, a compound loss function containing wavelet loss is proposed to facilitate the preservation of details. Also, an enhancement module with convolutional block attention is put forward to further refine the prediction quality. Experiments were conducted to compare the proposed ECPW-STFN with five state-of-the-art methods on the public CIA and Daxing datasets. The results show ECPW-STFN is better than GAN-STFM which also requires two images, and not inferior to the methods requiring at least three inputs, even exceeds the optimal MLFF-GAN in some cases, proving its great superiority and potential. The code will be available at https://github.com/lixinghua5540/ECPW-STFN.

Tomography of nonlinear materials via the monotonicity principle

In this paper we present a first non-iterative imaging method for nonlinear materials, based on Monotonicity Principle. Specifically, we deal with the inverse obstacle problem, where the aim is to retrieve a nonlinear anomaly embedded in linear known background. The Monotonicity Principle (MP) is a general property for various class of PDEs, that has recently generalized to nonlinear elliptic PDEs. Basically, it states a monotone relation between the point-wise value of the unknown material property and the boundary measurements. It is at the foundation of a class of non-iterative imaging methods, characterized by a very low execution time that makes them ideal candidates for real-time applications. In this work, we develop an inversion method that overcomes some of the peculiar difficulties in practical application of MP to imaging of nonlinear materials, preserving the feasibility for real-time applications. For the sake of clarity, we focus on a specific application, i.e. the Magnetostatic Permeability Tomography where the goal is retrieving the unknown (nonlinear) permeability by boundary measurements in DC operations. This choice is motivated by applications in the inspection of boxes and containers for security. Reconstructions from simulated data prove the effectiveness of the presented method.

Optimal Injection Attack Strategy for Nonlinear Cyber-Physical Systems Based on Iterative Learning

This paper aims to investigate the security problem of nonlinear cyber-physical systems (CPSs), which poses a challenge to handle compared with linear CPSs. A series of optimization problems for nonlinear CPSs under injection attack are constructed, which are based on a general model of the nonlinear systems with repetitive operation characteristics and a novel introduction of the key technical lemma. These optimization results are more general than the existing injection attack results and the requirements for attackers to obtain system information are relaxed. Also, the form of switching applied to the attack strategy possesses several advantages, including high stealthiness, lower cost, and more flexibility. Therefore, the new optimal injection attack strategies are expected to be more widespread and provide a basis for the design of defense strategies. The key to acquiring the designed optimal attack strategies is to adopt the linear input/output (I/O) data model for these systems via introducing an estimation term of the improved projection estimation method into the linear model. Finally, a networked GLUON-6L3 manipulator example validates the effectiveness of the proposed methods. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —The main purpose of this paper is to study the cyber security of nonlinear cyber-physical systems from the perspective of attackers, which can help defenders fully understand the behavior of attackers. Most of the existing attack strategies aimed at linear systems or have a priori knowledge of the attacked systems. However, there are great limitations and difficulties in practical application. In this paper, the new attack strategies are proposed by combining system identification, iterative learning and control theory, which relaxes the requirement of the attacker’s ability. In detail, the attacker only needs to obtain the control input and output data of the attacked system and design the attack strategy according to it. Among them, the first-in, first-out queue storage method is applied to remove the requirements for the storage capability of the attacker. In practical applications, the attacker does not need to store the I/O data continuously, but only store the initial value and the data of two successive iterations. The mathematical analytic forms of two optimal attack strategies are given, and then their effectiveness are verified on a welding work of a networked six-axis manipulator. In the future, we will investigate the design of attack strategies for nonlinear systems with heterogeneous dynamics and multiple delays.

Constraints and Countermeasures of Digital Economy Empowering High-quality Development in Chinese Agriculture

With the wide application of networking, informatization, and digitization technologies in agriculture, as well as the improvement of farmers’ modern information skills, the digital economy has become an important engine driving high-quality development and digital transformation in agriculture. Digital agriculture is an essential component of the national strategy for “Digital China” and an effective measure to achieve the deep integration of informatization and agricultural modernization. Driven by digital technologies, digital agriculture in rural areas shows tremendous development potential, huge market demand, and broad application prospects. The era of digital-driven agriculture and rural areas “leapfrogging” is approaching. However, the digital economy empowering high-quality development in agriculture also faces various constraints, such as lagging infrastructure construction, slow development of digital agriculture, insufficient endogenous driving forces, and difficulties in practical applications. To address these issues, the government should increase investment, clarify directions, promote industry academia research collaboration, and strengthen core technologies. These measures will enable the digital economy to better empower the high-quality development of Chinese agriculture.

A Hybrid Photorealistic Architecture Based on Generating Facial Features and Body Reshaping for Virtual Try-on Applications

&#x0D; &#x0D; &#x0D; Online shopping using virtual try-on technology is becoming popular and widely used for digital transformation because of sustainably sourced materials and enhancing customers’ experience. For practical applicability, the process is required for two main factors: (1) accuracy and reliability, and (2) the processing time. To meet the above requirements, we propose a state-of-the-art technique for generating a user’s visualization of model costumes using only a single user portrait and basic anthropometrics. To start, this research would summarize different methods of most virtual try-on clothes approaches, including (1) Interactive simulation between the 3D models, and (2) 2D Photorealistic Generation. In spite of successfully creating the visualization and feasibility, these approaches have to face issues of their efficiency and performance. Furthermore, the complexity of input requirements and the users’ experiments are leading to difficulties in practical application and future scalability. In this regard, our study combines (1) a head-swapping technique using a face alignment model for determining, segmenting, and swapping heads with only a pair of a source and a target image as inputs (2) a photorealistic body reshape pipeline for direct resizing user visualization, and (3) an adaptive skin color models for changing user’s skin, which ensures remaining the face structure and natural. The proposed technique was evaluated quantitatively and qualitatively using three types of datasets which include: (1) VoxCeleb2, (2) Datasets from Viettel collection, and (3) Users Testing to demonstrate its feasibility and efficiency when used in real-world applications&#x0D; &#x0D; &#x0D;

Ultra‐High DC and Low Impedance Output for Free‐Standing Triboelectric Nanogenerator

AbstractThe free‐standing triboelectric nanogenerator (F‐S TENG) is demonstrated to be an efficient energy conversion device. However, limited by the inherent features of high‐voltage, low alternating current and high output impedance, TENGs suffer from inefficient energy utilization and difficulties in practical applications. Therefore, the efficient harvest of mechanical energy and optimization of output are currently critical issues for TENGs. Here, a strategy is proposed for optimizing the inherent features by coupling F‐S TENG to a capacitor auto‐conversion circuit (CAC). For the first time, a specific idea is proposed explicitly to reduce the output impedance by the management circuit. Eventually, a high current density of 32 A m−2 and a charge density of 1.75 mC m−2 are achieved, which is 8.78 times higher than previous works. The power density and energy density reach to 124 W m−2 and 252 mJ m−2 under 470 and 150 Ω respectively. Then, 5000 LEDs and two lamps rated at 220 V and 50 W are manually illuminated. The resulting pulse output is utilized as a trigger source for the thyristor, enabling DC contactless switching. This work provides a novel approach to realize high‐performance DC output of TENGs and effectively reduce external load impedance.

Three-dimensional acoustic asymmetric focusing by mode-conversion structure

Acoustic asymmetric focusing (AAF) has attracted widespread attention owing to its great potential in medical ultrasound. The recent development of acoustic metamaterials and metasurfaces has provided various concepts to achieve AAF systems. Generally, the previously demonstrated AAF lenses were limited to one-dimensional structures, leading to difficulties in practical applications. The realization of AAF in three-dimensional (3D) space based on a two-dimensional (2D) device still remains a challenge. Here, we experimentally and numerically study a 2D AAF lens based on mode converters composed of two pairs of phased unit cells I and II and a step waveguide. Based on the phase profile of sound focusing and the step waveguide of mode converters, we experimentally design and demonstrate a type of 2D AAF lens and observe acoustic focusing and asymmetric transmission simultaneously in 3D space under the excitation of the zero-order wave. Both the phenomena arise from the phase modulation of the lens and the cutoff frequency of the first-order wave for the step waveguide, respectively. The fractional bandwidth (the ratio of the bandwidth to the center frequency) of the lens can reach about 0.14. The proposed 2D AAF lens, as well as its associated mode converter, paves the way for the asymmetric manipulation of sound in 3D space that has potential applications in practical sound devices.

Strategies for sensor virtual in-situ calibration in building energy system: Sensor evaluation and data-driven based methods

Sensor errors greatly affect the optimal control, safe operation, and energy efficiency of the building energy system. The sensor virtual in-situ calibration (VIC) method based on Bayesian inference can calibrate the sensor errors and does not need to remove existing sensors or add redundant sensors. In the previous study, the calibration of all sensors (non-differential calibration) and the physical model-driven based methods are applied in the simulation system. However, in the practical building energy system, these reduce calibration accuracy and efficiency. To solve these two negative factors, two calibration strategies are proposed: (1) sensor evaluation, and (2) data-driven based methods. Four cases are designed to demonstrate calibration accuracy and efficiency improvement by applying the calibration strategies in the variable air volume system. The results show that non-differential calibration causes difficulties in practical applications because of its low accuracy (38.10 %) and long calibration time (5136.35 s). The calibration accuracy of the physical model-driven based methods is 9.06 %, which completely deviates from the true value. The two calibration strategies are applied to improve the sensor calibration accuracy (up to 91.88 %) and reduce the calibration time (28.99 %) although it will increase some replacement costs. Meantime, the calibration accuracy of the temperature sensor is further improved (exceeding 94 %) by dividing the operating conditions according to the compressor power. In summary, the application of the calibration strategies can effectively overcome the negative impacts on calibration accuracy and efficiency, and increase the robustness of the VIC method in the practical building energy system.

Towards ambient temperature operation of Li metal batteries using UV-Crosslinked single-ion electrospun electrolytes

In spite of the fact that lithium metal batteries (LMBs) facilitate the diversification of energy storage technologies, their electrochemical reversibility and stability have long been constrained by side reactions and lithium dendrite problems. While single-ion conducting polymer electrolytes (SICPEs) possess unique advantages of suppressing Li dendrite growth, they deal with difficulties in practical applications due to their slow ion transport in general application scenarios at ∼25 °C. In this study, we develop novel bifunctional lithium salts with negative sulfonylimide (-SO2N(-)SO2-) anions mounted between two styrene reactive groups, which is capable of constructing 3D cross-linked networks with multiscale reticulated ion nanochannels, resulting in the uniform and rapid distribution of Li+ ions in the crosslinked electrolyte. To verify the feasibility of our strategy, we designed PVDF-HFP-based SICPEs and the obtained electrolyte exhibits high thermal stability, outstanding Li+ transference number (0.95), pleasing ionic conductivity (0.722 mS cm−1), and broad chemical window (greater than5.85 V) at ambient temperature. As a result of the electrolyte structural merits, the Li||LFP cells displayed excellent cycling stability (96.4% reversible capacities after 300 cycles at 0.2C) without additional auxiliary heating. This ingenious strategy is expected to providing a new perspective for advanced performance and high safety LMBs.

A combination of library search and Levenberg-Marquardt algorithm in optical scatterometry

Optical scatterometry has been widely used for nanostructure metrology in integrated circuit (IC) given the fast, low-cost, non-contact, non-destructive, and in-line technique. Library search is a common solution for optical scatterometry, but the deterministic error in this method is hard to be eliminated due to the limitation of the grid interval in the spectral library. Another solution is the nonlinear regression algorithm, which faces difficulty in practical applications in meeting the in-line real-time requirements as it involves the time-consuming rigorous coupled-wave analysis (RCWA) calculations in the forward modeling. In this paper, we propose a combination of library search and Levenberg-Marquardt (LSLM) algorithm to solve the problem. We apply a presorted K-dimensional tree (KD-tree) algorithm to speed up searching the initial parameters in the spectral library and replace the time-consuming RCWA calculations with the Taylor expansion in the spectral library to compute the Jacobian matrix in the Levenberg-Marquardt (LM) procedure. Simulations and experiments demonstrate that the results in the proposed method are accurate, and in particular, the speed is much faster than that in previous methods.

A simple but effective capacity model for check and design of beam-column joints in RC seismic buildings

The design of beam-column joints is a crucial step of the design of seismic-resistant RC framed structures. Premature joint failure may prevent yielding of beams and the achievement of the target collapse mechanism. Hence, an effective capacity model is essential. In the past, Italian building code in force before NTC (2008) required that a minimum amount of ties, equal to that used in the column, be provided in the joint, but it did not stipulate any verification of the joint. Eurocode 8 (2005) and subsequently Italian NTC (2008 and 2018), recognize that shear in beam-column joint is higher than in columns and recommend the explicit verification of the joint. The suggested capacity model is based on stress analysis and is very demanding in terms of joint ties. The new draft of Eurocode 8 (2020) provides a different capacity model, based on research carried on by Fardis (2021), which better reflects the state of art on this topic under both theoretical and experimental point of view. This paper examines the capacity model for shear resistance of joints adopted in Eurocode 8 draft, underlining its advantages but also some difficulties in practical application. Based on a parametric analysis of a wide set of RC joints, it is showed that it is possible to assume that the joint rebars are yielded when the joint attains its ultimate limit state. This assumption does not determine any significant loss of precision in evaluating the shear resistance of the joint and leads to a capacity model consistent with other models adopted in Eurocode 8, e.g. those for shear resistance of RC beams. Furthermore, it allows a closed form solution of the problem that leads to a very simple procedure both for the verification of joint resistance and for the optimization of the design of horizontal and vertical rebars of the joint.

Pt-Based Oxygen Reduction Reaction Catalysts in Proton Exchange Membrane Fuel Cells: Controllable Preparation and Structural Design of Catalytic Layer.

Proton exchange membrane fuel cells (PEMFCs) have attracted extensive attention because of their high efficiency, environmental friendliness, and lack of noise pollution. However, PEMFCs still face many difficulties in practical application, such as insufficient power density, high cost, and poor durability. The main reason for these difficulties is the slow oxygen reduction reaction (ORR) on the cathode due to the insufficient stability and catalytic activity of the catalyst. Therefore, it is very important to develop advanced platinum (Pt)-based catalysts to realize low Pt loads and long-term operation of membrane electrode assembly (MEA) modules to improve the performance of PEMFC. At present, the research on PEMFC has mainly been focused on two areas: Pt-based catalysts and the structural design of catalytic layers. This review focused on the latest research progress of the controllable preparation of Pt-based ORR catalysts and structural design of catalytic layers in PEMFC. Firstly, the design principle of advanced Pt-based catalysts was introduced. Secondly, the controllable preparation of catalyst structure, morphology, composition and support, and their influence on catalytic activity of ORR and overall performance of PEMFC, were discussed. Thirdly, the effects of optimizing the structure of the catalytic layer (CL) on the performance of MEA were analyzed. Finally, the challenges and prospects of Pt-based catalysts and catalytic layer design were discussed.

Improving diagnosis accuracy of non-small cell lung carcinoma on noisy data by adaptive group lasso regularized multinomial regression

Non-small cell lung carcinoma (NSCLC) accounts for 80–85% of all lung cancers, ranking first in the cause of death of malignant tumors. Early and accurate diagnosis of NSCLC is crucial for the follow-up treatment, which is helpful in improving the survival time and quality of life of patients. Machine learning methods based on gene expression profile data provide a new way for the early diagnosis of NSCLC. However, there are a few difficulties in practical application: data noise, overlapping gene grouping, and the importance evaluations for gene groups and individual genes. This paper aimed to deal with these problems by developing an adaptive group lasso regularized multinomial regression (AGLRMR). Robust principal component analysis (RPCA) was presented to separate clean data from noisy gene expression profile data of NSCLC. Weighted gene co-expression network analysis (WGCNA) was adopted to perform overlapping gene grouping. The importance evaluation criteria for individual genes and gene groups were proposed by fusing module membership, information theory, and symmetric uncertainty. Compared with RAMRSGL, AMRSOGL, MSGL, MRGL, and MR-lasso, AGLRMR improved the diagnosis accuracy of NSCLC by 1.2%, 2.3%, 2.2%, 2.5%, and 4.0% on NSCLC-1 dataset, and by 1.1%, 8.7%, 6.2%, 7.1%, and 16.3% on NSCLC-2 dataset, respectively. The operation speed of AGLRMR is consistent with AMRSOGL and slower than the other four methods. Biological analysis showed that the genes selected by AGLRMR were highly correlated with NSCLC.

SECRET IMAGE SHARING AND IDENTIFYING FAKE IMAGE USING BLOCK CHAIN

The detection of fraudulent images and the attribution of image sources has become a hot topic of debate in both the news industry and social media. Anyone may quickly create or edit digital content in the age of digitalization, and they can then simply broadcast it on social media networks. The use of these social networking sites has created new difficulties for practical application, such as the viral transmission of incorrect or misleading material with harmful intentions. On the one hand, they greatly facilitate modern communication. A currently very emerging technology for data sharing and application is block chain. By using data encryption, timestamps, and distributed consensus, it is possible to exchange decentralised information in distributed systems without relying on one another, which increases the effectiveness of data sharing and application. The massive data remote sensing image system can fully exploit this technology, and the multi-system shared node storage system can be controlled effectively and uniformly to increase the system's economic efficiency. This study suggests important research technologies, builds the shared architecture based on block chain technology, and offers a theoretical framework for non-engineering practise.

Magnetic Tile Surface Defect Detection Methodology Based on Self-Attention and Self-Supervised Learning.

As the core component of permanent magnet motor, the magnetic tile defects seriously affect the quality of industrial motor. Automatic recognition of the surface defects of the magnetic tile is a difficult job since the patterns of the defects are complex and diverse. The existing defect recognition methods result in difficulty in practical application due to the complicated system structure and the low accuracy of the image segmentation and the target detection for the diversity of the defect patterns. A self-supervised learning (SSL) method, which benefits from its nonlinear feature extraction performance, is proposed in this study to improve the existing approaches. We proposed an efficient multihead self-attention method, which can automatically locate single or multiple defect areas of magnetic tile and extract features of the magnetic tile defects. We also designed an accurate full-connection classifier, which can accurately classify different defects of magnetic tile defects. A knowledge distillation process without labeling is proposed, which simplifies the self-supervised training process. The process of our method is as follows. A feature extraction model consists of standard vision transformer (ViT) backbone, which is trained by contrast learning without labeled dataset that is used to extract global and local features from the input magnetic tile images. Then, we use a full-connection neural network, which is trained by using labeled dataset to classify the known defect types. Finally, we combined the feature extraction model and defect classification model together to form a relatively simple integrated system. The public magnetic tile surface defect dataset, which holds 5 defect categories and 1 nondefect category, is used in the process of training, validating, and testing. We also use online data augmentation techs to increase training samples to make the model converge and achieve high classification accuracy. The experimental results show that the features extracted by the SSL method can get richer and more detailed features than the supervised learning model gets. The composite model reaches to a high testing accuracy of 98.3%, and gains relatively strong robustness and good generalization ability.

Biological Invasions in Neotropical Regions: Continental Ichthyofauna and Risk Assessment Protocols.

The objectives of this study were to compare four risk assessment protocols for non-native species in neotropical regions and to assess the potential application of these tools for the management of invasive species and conservation of the ichthyofauna in Brazil. The protocols Fish Invasiveness Screening Kit (FISK), Aquatic Species Invasiveness Screening Kit (AS-ISK), European Non-Native Species in Aquaculture Risk Assessment Scheme (ENSARS) and Fish Invasiveness Screening Test (FIST) were applied for the species Oreochromis niloticus, Coptodon rendalli, Poecilia reticulata and Apteronotus aff. albifrons, whose sources of introduction are aquaculture and fishkeeping. The species were classified as low, medium or high risk of invasion. The scores of the species O. niloticus, C. rendalli and P. reticulata classified them as high risk of invasion in all protocols, whereas A. aff. albifrons had medium risk in the protocols FISK and AS-ISK and low risk in the FIST. Although the results were similar for species whose impacts are widely described, less studied species may have their classification compromised by the lack of evidences in the literature. Despite the difficulties for practical application, the use of these tools may be encouraged, considering the potential threats of other invasive species emerging in Brazil. The comparison between the methods showed that the use of AS-ISK, combined with ENSARS in cases of introductions by aquaculture, provides important answers about ecological impacts on natural environments and about the stages of the aquaculture production chain that should be better inspected.

Research on the Key Technology of Web Data Extraction and Mining Based on the Probability Distribution

In the development of internet technology innovation, the quantity of big data information continues to rise, showing the characteristics of dynamic, heterogeneous, massive, and so on. How to explore valuable and potential knowledge from the network system is the main topic of research and scholars. Based on the understanding of web data extraction and web mining technology, this paper makes a systematic study of web text mining, proposes the most commonly used hidden Markov model in probability distribution, and constructs the corresponding text mining method. The hidden Markov model is a statistical model, which is mainly used to represent a Markov process with unknown parameters. The difficulty in practical application lies in how to define the hidden parameters of the process from observable parameters and then use these parameters to conduct in-depth research. Finally, the practical results show that the hidden Markov model can not only obtain the information of different regions but also deeply analyze the data set, which proves the feasibility of this research technology.

State-updating-based DOA estimation using sparse Bayesian learning

Recent studies of the direction-of-arrival (DOA) estimation reveal that the methods based on sparse Bayesian learning (SBL) exhibit many advantages over conventional approaches. However, these methods still face difficulties in practical applications due to their high computational complexity. To address the problem, a state-updating-based DOA estimation method using sparse Bayesian learning is proposed in this paper. In the proposed method, the state filter is employed to establish a recursive link among the source states within the observation time, and a probability distribution-based fitting technique is developed to fit the initial values for the iterative process by the source states. Numerical simulations and experimental results demonstrate that, the proposed method yields significantly reduced computational complexity and improved DOA estimation accuracy.

Advanced Retinex-Net image enhancement method based on value component processing

When capturing images under low-light lighting conditions, the resulting images often suffer low visibility. Such low-visibility images not only affect the visual effect but also cause many difficulties in practical application. Therefore, image enhancement technology under low-light conditions has always been a challenging problem in image algorithms. Considering that most of the existing image enhancement methods are based on the RGB color space enhancement technology, the correlation among the RGB three primary colors is ignored, which makes the color distortion phenomenon easy to occur when the image is enhanced. To solve the problems of poor image visibility and color deviation under low-light conditions, in this paper an advanced Retinex network enhancement method is proposed. In the method, firstly the low-light RGB image is transformed into HSV color space, the Retinex decomposition network is used to decompose and enhance the value component separately, and thus increasing the resolution of the value component through up-sampling operation; then, for the hue component and saturation component, the nearest neighbor interpolation is used to increase their resolutions, and the enhanced value component is combined to convert back to RGB color space to obtain the initial enhanced image; finally, the wavelet transform image fusion technology is used to fuse with the original low-light image to eliminate the over-enhanced part in the initial enhanced image. The analysis of experimental results shows that the method proposed in this paper has obvious advantages in brightness enhancement and color restoration of low-light images. Especially, comparing with the original Retinex network method, the NIQE value decreases by an average of 19.49%, and the image standard deviation increases by an average of 41.35%. The algorithm proposed in this paper is expected to be effectively used in the fields of security monitoring and biomedicine.