13 results about "Rain removal" patented technology

The application discloses a low-illumination rain-added image enhancement method based on a PERM-Net (Physics-guided Event-driven Rain Modulation Network), and proposes a joint enhancement network integrating degradation prior, event-driven coding and state space recovery in view of the problems of serious image brightness attenuation, obvious continuous rain streak interference, texture detail loss and insufficient perception ability of the enhanced result. The method comprises the following steps: 1) improving an existing data set; 2) constructing a degradation prior estimation module to jointly model the brightness attenuation, rain streak density and direction structure information in the image; 3) designing an event rain streak coding module to dynamically code the continuous rain streak features and extract the direction structure; 4) constructing a physics-guided rain streak Mamba block to realize image brightness recovery, rain streak suppression and global structure reconstruction; and 5) combining an adaptive fusion module and a perception optimization head to improve the detail expression ability of the enhanced image. The method realizes excellent rain removal effect and can generate an image rich in details and visually realistic in a complex environment.

The application discloses a single image rain removal method based on cyclic inner and outer dense connection and attention enhancement, comprising the following steps: constructing a cyclic inner and outer dense connection and attention enhancement network for single image rain removal, wherein the network comprises a cyclic subnetwork with multiple shared parameters; determining a loss function, wherein the loss function comprises a structural similarity and a perception loss; inputting training data; and training network parameters to obtain a cyclic rain removal model. The application establishes dense connection in the cycle, expands the channel of network information transmission, makes the shallow layer features containing more background detail information fully utilized by the deep layer and the subsequent convolution layer of the network through information transmission, and enables the network to retain more background details in the cyclic progressive rain removal process.

The application provides a video rain removal method and system based on a space-time difference state space model, and belongs to the technical field of deep learning. The application solves the problem of how to ensure accurate rain removal while retaining the background texture to the greatest extent and breaking the trade-off between "rain removal completeness" and "background fidelity". The method comprises the following steps: S1: extracting shallow space-time features from an input video; S2: constructing anchor points and reference features by using multi-plane scanning, generating a gating map by calculating geometric consistency, and removing rain information in the features by using a difference mechanism; S3: during training, capturing local space-time details by using orthogonal plane convolution, during inference, enhancing the video background texture by using a single-stream convolution after reparameterization, and outputting a rain-removed video. The system comprises a 3D convolution layer module, a consistency-guided difference gating module, and a reparameterized three-plane feedforward network.

The application discloses an image rain removal method based on learnable prior distillation and frequency domain cross attention, and relates to the field of image processing, comprising: based on rainy image data, performing degradation analysis on the rainy image data through a prior representation network to extract a rain streak degradation mode, wherein the rain streak degradation mode contains prior knowledge; inputting the rainy image data into a pre-trained teacher network to extract teacher knowledge; constructing a student model based on a cross attention mechanism, taking the prior knowledge as a query, taking the teacher knowledge as a key and a value to perform prior distillation, and generating prior features; introducing a frequency domain feature extraction module to perform spatiotemporal feature fusion on the rainy image data to generate fused features; and performing image reconstruction and recovery based on the prior features and the fused features to generate clear image data. The application solves the technical problems of poor rain streak removal effect and limited model practicability of existing single-image rain removal technology, and achieves the technical effects of accurate rain streak separation and enhanced model generalization ability.

The application discloses a space-frequency domain double-path coupling adaptive image rain removal method, and solves the technical problems of excessive dependence on spatial domain features, insufficient mining of frequency domain discriminative information in the prior art, low rain removal precision, easy loss of details, and weak generalization ability in complex rain conditions. The method comprises the following steps: constructing a double-domain collaborative block, building a rain removal network with the double-domain collaborative block as the core, extracting shallow features, extracting space-frequency domain double-path parallel features, adaptively fusing double-domain features, generating a rain removal residual image, and reconstructing a clear rain removal image. The application constructs a double-domain collaborative block, and further constructs a rain removal network with the double-domain collaborative block as the core. Through space-frequency domain double-path parallel modeling, the application realizes adaptive weighted fusion of double-domain features in combination with difference perception attention, effectively decouples rain streaks and background details, improves the image restoration quality in complex rain conditions, and is applied to intelligent monitoring, automatic driving, unmanned aerial vehicle aerial photography, remote sensing imaging, and rain day vision enhancement scenes.

The application discloses a low-illumination and rain superposition image enhancement method based on a TSDFF-Net, and comprises the following steps: 1) constructing LLR-Train and LLR-Test data sets; 2) training a rain prior feature extraction module; 3) training a low-illumination enhancement module; and 4) testing the TSDFF-Net. According to the method, the model enhancement quality can be improved according to rain mark information, and the method has good effects on rain removal and low-illumination enhancement of a weak-light rainy-day image.

The present application relates to a kind of image rain removal enhancement method and system based on agent planning, belong to image processing technical field.The present application is to solve the problem of insufficient flexibility of existing rain removal technology in processing real world complex rain degradation.Method includes: S1: the preliminary rain removal result of the output of basic rain removal model is input into shared backbone network and is degraded to perceive and enhance strategy planning;S2: discrete path P in tool scheduling module is called, and preliminary rain removal result is handled by spatial self-adaptive successive enhancement with intensity scheduling module, and image rain removal enhancement result is output.System includes: shared backbone network, tool scheduling module, intensity scheduling module and instance repair program.

The application discloses a video rain removal method based on an adaptive tensor weighted kernel norm, belongs to the field of computer vision and image processing, and comprises the following steps: acquiring a rainy video to be removed from rain lines; analyzing prior information with distinguishability of a rain-free video and rain lines; constructing a rain removal model for removing the rain lines in the rainy video based on the prior information; and removing the rain lines from the rainy video based on the rain removal model. The application can make the video after rain removal retain more detailed features, avoid rain line blurring, effectively avoid loss of video detailed information, make the video after rain removal retain more abundant and fine detailed features, and realize high-fidelity video rain removal.

The application discloses a multi-scale driven space-frequency dual-domain adaptive gating fusion image rain removal method, and belongs to the technical field of image restoration. In view of the shortcomings of the existing method in the complex rain scene, a space-frequency dual-domain fusion module is introduced in the multi-scale coding and decoding backbone structure, the space domain and frequency domain features are modeled on different resolution levels at the same time, the dynamic weighted fusion of the dual-domain information is further realized through the adaptive gating fusion mechanism, the expression ability of the model to the complex rain structure and the recovery ability of the model to the background details are enhanced, the frequency domain selection module is introduced to screen and enhance the spectral components of the features, and the fusion ratio between the space domain features and the frequency domain features is dynamically adjusted through the adaptive gating fusion mechanism, so that the model adaptively selects a better feature representation mode according to the texture complexity and structure distribution of different regions; the interactive optimization of the rain branch and the background branch is realized through the coupling representation module, and the residual rain and the background damage problem are reduced.

The application provides a wiper automatic control method, device, equipment and readable storage medium. The wiper automatic control method comprises the following steps: acquiring a raindrop image in real time, wherein the raindrop in the raindrop image is not in contact with the front windshield; calculating a total volume of raindrops in a preset range according to the raindrop image, averaging the total volumes of raindrops corresponding to all raindrop images acquired in a preset time to obtain a volume reference value; and controlling the working frequency of the wiper according to the volume reference value. The application expands the sensing range of raindrops through the image acquisition mode, which helps to reduce misjudgment and missed judgment. The volume reference value is calculated to determine the rainfall size, and the judgment accuracy is high, so that the stable and good rain removal effect is ensured, and the driving comfort and safety are improved.

The receptive field enhancement and multi-attention mechanism integrated image rain removal method belongs to the field of automatic driving or intelligent driving environment perception. The method of the present application is as follows: feature extraction is performed on the input image containing rain, a convolution feature map with a multi-scale receptive field is constructed to enhance the network's perception ability of rain streaks of different scales; on this basis, parallel channel attention, spatial attention and residual attention mechanisms are introduced to fully exploit the correlation between features and improve the ability to distinguish complex rain streaks; the global modeling capability and local enhanced features are integrated in the Transformer structure to realize joint modeling of global and local rain streak information of the image; a training set containing real rain images and rain-free images is constructed, and the network weight is optimized in a supervised learning manner; in actual application, the image containing rain is input into the trained model, and a clear rain-removed image is output. The present application can remove rain streaks of different intensities and different morphologies while maintaining image details and texture features.

This application relates to underwater image enhancement methods, devices, equipment, and media. The method, through innovative design, implements a progressive image inpainting structure with three stages: a Transformer encoder, U-net, and FSnet, and the TCFF and CSFF mechanisms between these stages. This method achieves performance breakthroughs in three dimensions: global semantic information modeling, local semantic information extraction, and spatial detail preservation. It exhibits superior performance in these three dimensions. Based on a Transformer and CNN fusion architecture, this method also possesses high inference efficiency, making it valuable for underwater robot operations that meet the requirements of high-precision image enhancement and real-time processing. Balanced optimization in multi-dimensional semantic representation and detail preservation improves the model's generalization ability, making it applicable to various image inpainting tasks such as rain removal, defogging, and deblurring.