236 results about "Image gradient" patented technology

The invention relates to the field of image processing, in particular to a seamless steel tube defect detection method based on image processing, which comprises the following steps: acquiring an image of a sized steel tube through a high-resolution industrial camera, and preprocessing to enhance the image quality; designing an adaptive filter based on image gradient direction and spatial frequency characteristics, and enhancing defect characteristics; extracting a defect region by adopting adaptive threshold segmentation and morphological operation, and constructing a comprehensive feature vector through a Hu moment and a gray-level co-occurrence matrix; and finally, defect identification is realized by using a support vector machine classifier, and technical support is provided for steel pipe production quality control. According to the method, the adaptive filter is constructed based on the gradient direction and the spatial frequency, and efficient detection and classification of seamless steel pipe surface defects are realized by combining the comprehensive feature vector and the support vector machine classifier.

The invention relates to the technical field of medical image analysis, in particular to a CAR-T cell culture monitoring system based on image recognition, which comprises a cell region division module, a cell morphology analysis module, a cell behavior dynamic analysis module and a cell population collaborative monitoring module. According to the method, through cell image gradient intensity analysis, refined cell edge detection and dynamic adjustment of a segmentation threshold value, high-precision extraction of cell boundaries is ensured, noise interference is avoided, cell shapes, textures and geometric features are refined, the identification degree of morphological features is improved, and tiny changes of cell morphologies are accurately captured; dynamic tracking and trend evaluation are carried out on cell behaviors based on adjacent frame images, tiny dynamic changes such as cell division, aggregation and migration are accurately reflected, excessive smoothness in the dynamic process is avoided, cell population behavior monitoring is combined with interaction and corresponding speed calculation, the population synergistic effect change trend is comprehensively evaluated, and the cell population behavior monitoring effect is improved. And finer cell culture environment optimization and clinical research support are provided.

The invention relates to the technical field of image processing, in particular to a super-resolution reconstruction system and method based on a pulse neural network, and the system comprises an image pulse encoder, a pulse feature enhancer and a differentiable pulse decoder. An image pulse encoder simulates a receptive field of a retina by using DoG response, performs adaptive pulse distribution on an input low-resolution image in combination with an image gradient, and converts the low-resolution image into a space-time pulse sequence reflecting a high-frequency region and a low-frequency region in the low-resolution image; a pulse feature intensifier performs coarse-grained and fine-grained structure reconstruction of a low-resolution image on the space-time pulse sequence by using a pulse time sequence dependent plasticity mechanism to obtain a granularity feature pulse; and the differentiable pulse decoder converts the granularity characteristic pulse to obtain a corresponding high-resolution image. According to the method, efficient and low-consumption image super-resolution reconstruction is realized through a bionic retina coding mechanism and a pulse time sequence optimization strategy.

The invention discloses a multi-modal remote sensing image matching method, particularly relates to the technical field of remote sensing image processing, and is used for solving the problem of multi-modal image matching. The method mainly comprises the following steps of: 1, improving a phase consistency model, and constructing a phase-moment weighted joint direction feature in combination with a maximum moment and a minimum moment to replace the feature expression of the traditional image gradient; 2, implementing a point product fusion strategy on the phase-amplitude characteristics extracted by the phase consistency model and the maximum moment, and constructing phase-moment weighted joint amplitude characteristics; and 3, on the basis of the steps 1 and 2, identifying the direction of the feature points and screening local peak values to determine the main direction. Three values adjacent to a peak value are selected, and the peak value position is interpolated through parabola fitting so as to improve the matching precision; and step 4, constructing a logarithm polar coordinate descriptor based on regularization non-uniform partition to generate a feature description vector. Through the mode, high-precision and high-efficiency matching of the multi-mode remote sensing image can be realized.

The invention relates to the technical field of image precision optimization, in particular to a surveying and mapping image mathematical precision evaluation method based on dense point cloud matching, which comprises the following steps: acquiring and preprocessing image data, synthesizing an image gradient modulus length and a maximum principal curvature, and introducing a surface normal vector partial derivative to obtain a surface normal vector partial derivative; calculating a curvature-driven depth discretization step length, and generating an initial point cloud data set; constructing a point cloud complex network structure diagram and calculating the topology durability of a point cloud topology structure to obtain an optimized point cloud data set; based on the optimized point cloud data set, in combination with the target image information, the adaptive curvature gradient weight and the error diffusion control value, constructing an adaptive matching cost function, and calculating a point cloud matching cost matrix; and calculating an optimal point cloud transformation matrix based on the point cloud matching cost matrix. According to the surveying and mapping image mathematical precision evaluation method based on dense point cloud matching, low-texture region point cloud matching error suppression and structure optimization of high-curvature region point cloud on a complex curved surface are realized.

Systems and methods for detecting anomalies in aviation data communication systems (e.g., air traffic control surveillance systems), include a processor receiving device status information. A variational autoencoder receives and optimizes the device status information and determines whether it qualifies as an anomaly. Optimized device status information is compared to either non-anomalous or anomalous device status data in a latent space of the variational autoencoder. The latent space preferably includes an n-D point scatter plot and hidden vector values. The processor optimizes the device status information by generating a plurality of probabilistic models of the device status information and determining which of the plurality of models is optimal. A game theoretic optimization is applied to the plurality of models, and the best model is used to generate the n-D point scatter plot in latent space. An image gradient sobel edge detector preprocesses the device status information prior to optimization.

The invention discloses a digital imaging method of an ear-nose-throat examination endoscope, and relates to the technical field of medical treatment, and the method comprises the following steps: synchronously collecting a white light image, a photoacoustic signal, a stimulated Raman spectrum, pressure sensor data and a physiological signal of a target orifice through a multi-physical field probe; constructing a four-dimensional tensor bound with the anatomical features; based on a vocal cord vibration fundamental frequency harmonic characteristic optimization graph convolutional network, dynamically constructing an adjacent matrix and coupling cross-modal characteristics to generate a submucosal lesion enhanced image; a deformable convolutional network constrained by vocal cord biomechanics is adopted, and the shape of a convolution kernel is dynamically adjusted according to the relation between real-time strain and elastic modulus, so that motion artifacts caused by swallowing actions are inhibited; generating a curvature-driven asymmetric convolution kernel based on ear canal spiral geometry, and executing super-resolution reconstruction in combination with confrontation training of fractal constraint; and fusing the white light image gradient and the pressure gradient field, and outputting a three-dimensional lesion contour consistent with the anatomical structure.

The invention provides an image multi-modal fusion method and system for a complex dynamic environment, and belongs to the technical field of multi-modal image processing. The method comprises the steps that a visible light image and an infrared image are acquired and preprocessed; respectively inputting the preprocessed images into a multi-layer encoder, and obtaining a comprehensive feature map of the images; inputting the comprehensive feature map into a bidirectional adapter, and converting the comprehensive feature map into two-dimensional features; dynamic alignment of the two-dimensional features is realized through dynamic adaptive adjustment, and then the two-dimensional features after dynamic alignment adjustment are restored to original dimensions; performing semantic alignment on the visible light image comprehensive feature map and the infrared image comprehensive feature map which are restored to the original dimensions; and calculating image gradient information after semantic alignment, determining a fusion weight, and fusing the comprehensive feature map by using the fusion weight to obtain a fused image. The capability of capturing multi-level information in a complex environment is enhanced, efficient fusion of infrared and visible light multi-mode images is achieved, and good robustness is achieved.

The invention provides a steel rail seam width detection method and system based on line-line distance, and the method comprises the steps: shooting an image of a rail connection part, obtaining a near-end rail and a far-end rail through an image recognition strategy, defining the boundary points of the top edge of the near-end rail as A and B, and defining the boundary points of the bottom edge of the near-end rail as C and D, the B and the C are located on the same track connecting line, the far-end track is defined in the same mode, and the A point and the B point are obtained; according to the C point determination strategy, determining the C point based on multi-factor weighting, and obtaining a near-end key point set; acquiring a far-end key point set according to the image acquisition processing step; two sets of key lines of the near-end track and the far-end track are obtained respectively, and the distance between the two sets of key lines is calculated to serve as the width of the steel rail seam; through multi-level constraints of image gradient features and track theoretical width, the problem that a single image feature is susceptible to noise interference is solved, and the positioning precision of A and B points is improved.

The invention discloses an image quality evaluation method based on dynamic weight feature extraction and three-dimensional fusion under a meta-learning framework, and the method comprises the steps: designing a multi-receptive-field and dynamic parameter module in an image quality evaluation model, extracting feature information under different views, and enriching and weighing multi-receptive-field features; screening out a salient region by using image gradient information outside the image quality evaluation model, designing a distortion classification model based on the salient region, and providing shared features for the image quality evaluation model; using random cutting and three-dimensional convolution to deeply fuse the multi-receptive-field features and the shared features to obtain deep fusion features; shartcut Vision Transform is proposed to perform semantic analysis on the fusion features, the association between the overall features and quality fluctuation is summarized, and the quality score of the image is obtained through linear mapping. The multi-distortion image quality evaluation method can accurately evaluate the quality of the multi-distortion image, can effectively process a large amount of data, and has the advantages of being objective, rapid and high in usability.

Network security anomaly detection systems and methods include a processor, in communication with the network, receiving network device status information. A variational autoencoder receives the device status information, optimizes the device status information, and determines whether the device status information qualifies as an anomaly. Optimized device status information is compared to either non-anomalous or anomalous device status data in a latent space of the variational autoencoder. The latent space preferably includes an n-D point scatter plot and hidden vector values. The processor optimizes the device status information by generating a plurality of probabilistic models of the device status information and determining which of the plurality of models is optimal. A game theoretic optimization is applied to the plurality of models, and the best model is used to generate the n-D point scatter plot in latent space. An image gradient sobel edge detector preprocesses the device status information prior to optimization.

The invention discloses a reversible information hiding method and system for enhancing image smoothness, and relates to the technical field of information security, and the system comprises a preprocessing module, an information embedding module and an optimization recovery module. According to the method, the Sobel operator is utilized to calculate the image gradient map, the region is divided based on the threshold T, the smooth region and the texture region can be accurately distinguished, an accurate basis is provided for differential processing of different regions, region characteristics are fitted, large-size blocks and sub-blocks of the smooth region are divided to facilitate subsequent refinement processing based on the gradient mean value, and the processing efficiency is improved. The small-size blocks of the texture area are beneficial to operation aiming at texture features, the rationality and effectiveness of overall processing are improved, meanwhile, the pixel mean value in the large-size blocks of the smooth area is calculated, the pixel and mean value difference value is recorded, reversible low-pass filtering is carried out, and the reversibility of the image in the whole information hiding and recovering process is ensured.

The invention belongs to the technical field of computers, and particularly relates to a reflection interference removing method and system based on reflection area detection and edge attention, and the method comprises the steps: employing the idea of a generative adversarial network on a main body, enabling a mixed image containing reflection interference to serve as the input of a model after preprocessing; inputting an input mixed image containing reflection interference into a reflection region detection network module, and obtaining multi-scale Gaussian-Laplacian features of the image through a reflection feature enhancement network module, and finally participating in synthesis of image gradient features of enhanced reflection; the obtained image gradient features serve as the input of the whole generation network, the features are extracted through the shallow feature extraction network, the extracted features are transmitted to the discrimination network in a jump connection mode, the output image after reflection interference removal is obtained, meanwhile, the quality of the generated image after reflection interference removal is discriminated, and the output image after reflection interference removal is obtained. And guiding the generation of the generator in the next round until the output quality meets the requirement. According to the invention, the influence of reflection interference on the image can be eliminated in a complex light external scene, and the quality of the license image acquired by the system is improved, so that the license image meets the image quality requirements of other downstream tasks.

The invention discloses a self-adaptive image fog removal method guided by a large-scale multi-modal model, belongs to the technical field of visual information processing, and is applicable to the field of foggy day scene image processing. The implementation method of the invention comprises the following steps: 1, obtaining degradation prior by using a pre-trained large-scale multi-modal model; 2, inputting the fog image and the degradation priori into a MoE-SSM model for training, so as to dynamically adjust model parameters and further perform dynamic perception defogging; 3, fusing the shallow feature Fs and the optimized feature # imgabs0 #, and reconstructing a clean and fogless image; 4, training a neural network by using a clean and fogless image and a truth value clear image by using a Charbonier loss function shown in a formula (8), so as to punish the deviation between the recovered image and the truth value clear image and encourage a consistent image gradient; compared with the prior art, the problem of compromise processing of receptive fields and calculation efficiency in existing defogging model image processing is solved, meanwhile, the requirement for a large amount of annotation data is reduced, and then the image processing accuracy and stability are improved.

A method for displaying and processing an image includes capturing the image, calculating one or more of a size, a position, and a field of view of a first region of interest (ROI) of the captured image, calculating one or more of a size, a position, and a field of view of a second ROI of the captured image, calculating a blur factor for blurring the first ROI, manipulating an image gradient of the first ROI pixel by pixel using the blur factor to reduce at least one of a brightness or a contrast of one or more pixels of the ROI, manipulating the second ROI by changing a color of one or more pixel values from the second ROI, and displaying the processed image on a display. An imaging system which includes an image signal processor (ISP) is also described.

The invention discloses an image processing method and system for extrusion shaping of a wood-plastic plate, and the method comprises the steps: carrying out the multi-scale analysis of an image of the wood-plastic plate, and recognizing the local features in the image; on the basis of multi-scale image feature extraction, a deformation mode in an image sequence is recognized through a time sequence image data association and dynamic correction method, and image processing parameters are adjusted in real time; according to the dynamically corrected image, processing the image by using morphological operation, and extracting a surface defect area of the wood-plastic plate; after morphological operation, identifying a defect area based on an image gradient optimization method; and on the basis of the sharpened image, analyzing the local flatness in the image by using a region growing algorithm. According to the method, the feature points are extracted from the image sequence and matched, the deformation mode between the images is recognized, and the local surface flatness is accurately analyzed in combination with the region growing algorithm, so that accurate detection of the surface defects and irregular regions of the wood-plastic plate is realized.

The disclosure deals with system and method for image gradient and derivative computation image processing. Noise, image sharpness, orientation, empirical parameters, and computational complexity are examples of image gradient and derivative computation challenges. Many traditional kernel-based operators excel at tackling one of these problems, but trade off their ability to handle others. Two new gradient detection kernels based on two-dimensional high order Taylor Series expansion tackle many such problems. The first kernel uses a wide range of the pixels in view to suppress noise, thereby improving the gradient intensities of edges. The second kernel builds on the first to leverage its noise suppression benefits while tackling an additional problem of degraded and low contrast edge boundaries. It can detect smooth lines in the presence of discontinuities and poor quality. The filter architecture allows for precise gradient calculation, edge detection, and orientation determination to less than one degree of the true value even when faced with signal to noise ratios that exceed 0.75.

The invention provides a fusion method for an infrared image and a visible light image, and belongs to the technical field of computer vision and artificial intelligence, and the method comprises the steps: S1, constructing an image fusion network model for generating a fusion image; s2, acquiring the infrared image and the corresponding visible light image to form a training data set, and training the fusion network model by using a loss function by using the training data set, the loss function is related to the structural similarity loss of the visible light image and the fusion image, the structural similarity loss of a semantic thermodynamic diagram mask area based on the infrared image and the corresponding visible light image, the image intensity loss and the image gradient loss; and S3, inputting an infrared image to be fused and a corresponding visible light image into the fusion network model trained in the step S2 to obtain a final fusion image. According to the method, manual operation is not needed, the computing power requirement is lowered, real-time image processing can be achieved, information is not prone to being lost, and efficiency is higher.

The invention provides an assembly type bridge structure damage identification method and system, and relates to the technical field of bridge structures, and the method specifically comprises the steps: firstly obtaining an image of an inter-beam connection point of an assembly segment, zooming the image to 224 * 224 pixels, and carrying out the graying processing; calculating an image gradient by using a Sobel operator, carrying out threshold edge detection, and extracting contour lines of connection points; preliminarily identifying the shape of the contour according to the circularity, the length-width ratio and the vertex number of the contour, fitting a circle by using a least square method, and calculating the circularity and dispersion of the contour; and finally, performing data processing and correlation analysis on the parameter change of the circular contour, generating a comprehensive damage index, and judging the damage type and degree of the connection point. The method can effectively identify and quantify the damage of the inter-beam connection point of the fabricated bridge segment in real time, and has high efficiency and accuracy.

The invention relates to the technical field of defect detection, in particular to a corrugated paper defect detection method and system based on machine vision. The method comprises the following steps: acquiring a corrugated paper image and converting the corrugated paper image into a frequency domain; identifying a periodic texture peak point based on amplitude spectrum angle energy distribution, generating a texture suppression filter, and combining with a multi-scale Gaussian high-pass filter to obtain a composite filter bank; the filter bank is used for filtering the frequency domain signals and performing energy weighted fusion; performing inverse transformation on the fused signal to obtain an enhanced image; and segmenting the enhanced image to obtain candidate defects, and verifying by combining the gradient information of the original image to determine a final defect. According to the scheme, the texture background can be adaptively inhibited, the multi-scale defect is enhanced, the false alarm rate is effectively reduced in combination with gradient check, and the detection accuracy and robustness are improved.

The invention discloses a visual calibration method, a terminal and a computer storage medium. The method comprises the following steps: generating theoretical grid coordinates; controlling a motion platform carrying a visual imaging unit to move to each target moving position in sequence, wherein the positions are obtained by superposing random two-dimensional offsets in a preset range; at each target moving position, a laser galvanometer and a visual imaging unit are synchronously triggered through the same hardware trigger source, and single-point mark engraving and image acquisition are executed; calculating a weighted centroid of the mark based on an image gradient magnitude, and extracting a sub-pixel center coordinate; combining the visual internal reference, the fixed coordinate transformation relation and the geometric constraint of the bearing plane, and resolving actually measured physical coordinates; carrying out space alignment, deviation calculation and invalid point interpolation on theoretical and actually measured coordinates, and generating a distortion data file which can be identified by a controller; the method further comprises a retry and fault-tolerant processing mechanism when coordinate calculation fails. According to the scheme, the calibration precision, the robustness and the automation level are effectively improved.

The invention relates to the technical field of image calibration, and provides a wafer defect automatic detection image calibration method, which comprises the following steps of: acquiring an initial image of the peripheral edge of a wafer by adopting an industrial line-scan digital camera; obtaining a grey-scale map of the wafer edge initial image, and calculating an image gradient value through convolution; performing threshold segmentation on the obtained gradient map, and extracting an edge feature region; sequentially carrying out image expansion and image corrosion processing on the edge feature region to obtain a processed edge feature region; calculating the area of the processed edge feature region, and extracting a feature curve; performing linear regression calculation on the characteristic curve to generate a mobile affine transformation matrix; and processing the initial image to obtain a calibration image. According to the invention, image dislocation and offset generated in an image acquisition process in a non-ideal environment can be corrected, the edge morphology of the wafer can be restored, and the image quality can be optimized. The invention further discloses a wafer defect automatic detection image calibration device and a computer readable storage medium.

The invention relates to the technical field of coating materials and detection, in particular to a surface defect detection method and system for a silicon carbide coating graphite disc. Performing multi-dimensional image data acquisition on the surface of the graphite disc; noise removal is carried out through an adaptive multi-scale convolutional neural network, and detail and edge information is retained in combination with image gradient constraints; high-frequency details are recovered by using an image enhancement technology, geometric correction is performed through a deep learning spatial transformation network, and data consistency is ensured; an authentication authority sequence is introduced, so that the authenticity and the generality of image data are ensured; an improved YOLOv9 model is adopted, an attention mechanism is introduced to enhance the extraction capability of target features, a Soft-NMS algorithm is combined to optimize a positioning frame, the overlapping detection problem is solved, and a traditional convolutional layer is replaced with an SCM module; a detection result is fed back in real time, a detailed detection report is generated, and decision support is provided. The surface defect detection precision and efficiency of the silicon carbide coating graphite disc are improved.

The invention relates to the field of image processing, in particular to a quality monitoring method and system for wire harness processing, and the method comprises the steps: obtaining a wire harness gray-scale map; for each pixel point of the grey-scale map, comparing a gradient histogram of a neighborhood of the pixel point with a preset template image gradient histogram, and finding out suspected matching points; calculating a gradient direction deviation angle between the suspected matching point and the template image, rotating a neighborhood of the suspected matching point based on the deviation angle, calculating a gray value of the suspected matching point after rotation, repeatedly rotating each pixel point in the neighborhood of the suspected matching point and calculating the gray value, and generating a corrected neighborhood; and matching the correction neighborhood of the pixel point with a preset template image, calculating a correlation coefficient between the correction neighborhood of the pixel point and the preset template image, and carrying out wiring harness abnormal region detection based on the correlation coefficient and a preset abnormal threshold value. According to the invention, by generating the correction neighborhood which is more fit with the template, the false detection and omission ratio can be effectively reduced, and the reliability of the detection result is improved.

A fissure occurrence parameter extraction method based on a borehole image relates to the technical field of geological resources and geological engineering, and mainly comprises the following steps: collecting and importing an original borehole image of a drilling area, enhancing and denoising the original borehole image, binarizing the image through dynamic area division, and obtaining a fissure occurrence parameter extraction result; detecting the edge of the fracture structural plane by using a Canny operator for carrying out adaptive threshold selection based on a borehole image gradient histogram, carrying out sine curve fitting on the fracture structural plane, and extracting fracture occurrence parameters on a fitted fracture sine curve; based on the engineering drilling image, the rock mass structural surface can be automatically identified, occurrence parameters can be extracted, previous manual intervention and operation are replaced, and the efficiency of analyzing geological conditions and underground water migration conditions in geological engineering is improved.

The invention relates to the technical field of image recognition control, in particular to a linear motion module accurate control method and system based on AI vision, and the method comprises the following steps: obtaining an interference fringe image and extracting a gray path, analyzing central point displacement to generate a coordinate sequence, repairing an interference feature position, and aligning the image and a control time sequence to generate a synchronization time table. The mapping structure response forms a set of control output configurations. According to the method, a peak path is constructed through interference fringe gray scale information in an image frame, central point pixel displacement is extracted, a coordinate migration sequence is constructed, fine-grained capture and tracking of an execution state are realized, an external interference area is identified according to image gradient change, vector repair is carried out, and image data stability and feature reliability are guaranteed. And aligning the repaired image response with a control instruction time sequence, constructing an impulse response lag statistical mechanism, realizing accurate mapping adjustment of the control frequency, and improving the control precision and response consistency in high-frequency displacement execution.

The invention discloses a microscopic visual threshold segmentation algorithm based on OTSU improvement for a micro-nano object image, and the algorithm comprises the steps: S1, combining image gradient information, calculating a new gray scale weight factor based on an index and a logarithmic function, and carrying out the bilateral filtering of an image; s2, calculating a traditional gray threshold value of the filtered image by using a traditional between-class variance formula, and calculating an improved gray threshold value by using an improved between-class variance formula based on a natural constant function; and S3, carrying out adaptive iterative operation on the pixels of the image by using the traditional gray threshold and the improved gray threshold so as to complete three-region threshold segmentation. According to the OTSU-based improved micro-nano microscopic visual image threshold segmentation algorithm disclosed by the invention, the problem of inaccurate calculation result caused by pixel quantity proportion imbalance during threshold segmentation is solved, and the limitation of poor local analysis capability of a global threshold segmentation algorithm is improved.

The invention discloses a multi-view vision hole site measuring method and device for a structural part, and relates to the technical field of multi-view vision. In order to solve the technical problems that an existing hole site detection method is insufficient in detection precision, low in detection efficiency, incapable of achieving comprehensive detection and unstable, the technical scheme provided by the invention includes that the multi-view vision hole site measurement method for the structural component comprises the following steps: loading a feature template image of a workpiece hole site, calculating similarity by using an image gradient field, and positioning the central position of a hole site from the workpiece image; a step of calibrating the position offset of the camera by using the correction target; according to the camera position, the light source condition and the image data of the hole site, the accurate coordinate position of the hole site of the workpiece in the three-dimensional space is calculated through a three-dimensional reconstruction method; and converting the three-dimensional coordinate data in the equipment coordinate system into a workpiece coordinate system. The method is suitable for being applied to the work of multi-view vision hole site measurement of the structural member.

The invention discloses a desulfurization optimization control method and system based on a multi-modal model, and relates to the technical field of industrial intelligent control, and the method comprises the steps: constructing a gas-liquid interface boundary point set, and generating a control feature vector; calculating fractal dimensions of control feature vectors by using a box counting method, constructing a disturbance point set, multiplying disturbance intensity by an expansion trend to obtain an unstable diffusion grade, matching key propagation points with nearest disturbance points by using nearest neighbor matching, and generating a control priority sequence; and defining a control unstable diffusion level, defining an optimization objective function, and solving by using constraint optimization to obtain a final control variable. According to the method, a composite strength factor is constructed through an image gradient field and a signal response ratio, a gas-liquid interface boundary point set and a control feature vector are constructed, the robustness and information density of features are improved, a main resonance frequency is recognized by combining a fractal interference field and short-time Fourier transform, key propagation points are screened by using wavelet transform and information entropy analysis, and a high-precision gas-liquid interface is obtained. The desulfurization efficiency and the stability of the reaction tower are improved.

The invention relates to a double-branch gradient information guided dark light image enhancement method, which comprises the following steps of: firstly, extracting low-illumination image gradient information by using a Sobel operator; a shallow network feature used for optimizing global brightness balance, a middle network feature used for correcting local area brightness of the image and a deep network feature used for carrying out contrast adjustment of scene perception in high-level semantic features are sequentially obtained through a depth feature enhancement branch; according to the image features after gradient information extraction, gradient features of a plurality of stages are obtained in sequence in combination with a gating gradient optimization module; and obtaining an enhanced image through an attention fusion module. According to the method, the problems of high-frequency detail loss and insufficient real scene generalization ability in a low-light environment in an existing method can be solved. Meanwhile, the enhanced image processed by the image enhancement deep network guided by the double-branch gradient information is not easy to have an overexposure phenomenon, and the feature description force of the network on a dark light image is effectively improved.