35 results about "Texture gradient" patented technology

The invention discloses a metal product defect detection method and system based on image recognition. The method comprises the following steps: firstly, executing reflection disturbance digestion processing on a surface image of a to-be-detected metal product to obtain a reflection digestion image; obtaining surface reference texture features of the defect-free metal product, and generating a reference texture model on the basis of a texture distribution rule, a gray average value and texture continuity; performing defect texture gradient separation on the reflection resolution image based on a reference texture model, positioning an abnormal region, and performing segmentation to obtain a suspected defect texture region; performing boundary pixel reconstruction on the suspected defect texture region to obtain a defect texture reconstruction region; obtaining a target defect texture region through local variance enhancement and neighborhood correlation analysis; and finally, the overexposure area is positioned, gray inverse stretching processing is performed, abnormal pixel group feature clustering is performed on the preprocessed defect area, and then a surface defect detection result is output, so that the precision and accuracy of metal product surface defect detection are improved, and the false detection rate is reduced.

The invention discloses a blast furnace burden surface segmentation method and device based on dynamic radius and multi-modal characteristics. The method comprises the following steps: initializing center point coordinates and radius of a blast furnace burden surface infrared image; dividing the blast furnace burden surface into a core high-temperature area, a relative high-temperature area and a low-temperature area based on the double temperature thresholds; establishing a radius double-constraint condition, and performing radius optimization under the condition that the condition is met; fusing the temperature feature and the texture gradient feature, calculating a multi-modal feature weight, and updating a centroid coordinate based on weighted average; executing a closed-loop feedback process of radius-centroid alternating optimization, and iterating until the central point and the radius converge; segmenting the image into a high-temperature central area and a low-temperature peripheral area based on the converged optimal parameter; differentiated processing algorithms are adopted for different areas, and Gaussian weighted fusion is adopted to obtain a globally consistent final result. According to the method, robust and self-adaptive region segmentation of the blast furnace burden surface is realized, a basis is provided for subsequent regional three-dimensional reconstruction, and the method has certain applicability and reliability.

The invention belongs to the technical field of image processing, and particularly relates to a silicate fireproof plate surface defect detection method based on machine vision, which comprises the following steps of: firstly, respectively acquiring a luminosity image representing material albedo and a morphology image representing surface microcosmic fluctuation by utilizing a line-scan digital camera in cooperation with a dual-channel light source; then, a micro texture index is constructed in combination with the gradient magnitude and the gray scale deviation, and a defect probability graph for inhibiting mineral interference is generated; and finally, through connected domain segmentation, calculating a penetration diffusion confidence coefficient by using an accumulated feature of an edge texture gradient, and verifying a suspected region. According to the method, the difference between the permeability characteristic of a real defect and the physical truncation characteristic of a natural mineral is utilized, the accuracy of distinguishing the permeability characteristic of the real defect and the physical truncation characteristic of the natural mineral in gray scale is improved, the false alarm rate is reduced, and the detection precision is improved.

This application discloses a method and system for detecting defects in metal products based on image recognition. The method first performs reflection perturbation elimination processing on the surface image of the metal product to be detected to obtain a reflection elimination image; then, it acquires the surface reference texture features of the defect-free metal product, and generates a reference texture model based on texture distribution patterns, gray-level mean, and texture continuity; next, it performs defect texture gradient separation on the reflection elimination image based on the reference texture model, locates abnormal regions, and segments them to obtain suspected defect texture regions; then, it performs boundary pixel reconstruction on the suspected defect texture regions to obtain defect texture reconstruction regions; the target defect texture region is obtained through local variance enhancement and neighborhood correlation analysis; finally, it locates overexposed regions and performs gray-level inverse stretching processing, and outputs the surface defect detection results after clustering abnormal pixel group features of the preprocessed defect regions, thereby improving the accuracy and precision of surface defect detection of metal products and reducing the false detection rate.

This invention provides a visual biomimetic edge detection method based on texture gradient modulation, belonging to the field of image edge detection technology. It focuses on introducing texture gradients to detect significant edges and suppress texture. First, the retina is modeled, and the image is encoded to obtain multiple information channels. Based on this, a novel peripheral modulation mechanism is designed, including unidirectional facilitation, lateral inhibition, and omnidirectional inhibition, to modulate the responses of simple cells in the primary visual cortex. Simultaneously, texture gradients are extracted using texture information and combined with the responses of simple cells to highlight the boundaries of textured regions and weaken the responses at textured edges. Next, endpoint cells in the second-level visual cortex are modeled to further modulate edge responses. Finally, the fourth-level visual cortex is modeled to integrate edge cues from all information channels to obtain the final edge detection result. This invention solves the problems of poorly highlighted textured region boundaries and significant noise in existing edge detection methods.

The invention provides an industrial vision intelligent online detection system for texture flaws of fiber products, which relates to the technical field of industrial machine vision and is characterized in that a spatial transformation network is introduced, differentiable geometric adaptive calibration is carried out on continuously acquired image streams, distorted fabric textures are mapped to a standard space, and the texture flaws of the fiber products are detected. And interference of physical deformation on detection is effectively eliminated. On this basis, a flawless ideal texture reference is generated by using a random block mask and pyramid feature reconstruction mechanism, and gradient sensing type adaptive difference calculation is performed on the calibrated image and the reconstructed image in combination with a texture gradient field of a reconstruction domain. And finally, accurately restoring the abnormal thermodynamic diagram of the calibration domain to a physical coordinate system through a reverse re-projection technology of a transformation grid, thereby realizing high-precision and anti-interference online detection of fine texture flaws on the surface of the fiber product in a dynamic deformation environment.

The invention relates to the technical field of coal seam image processing, and particularly discloses a coal seam separation boundary extraction method based on deep learning, which is used for solving the problem that pseudo-edges, fractures and horizon drifts easily occur in interface segmentation and edge extraction due to interference of water film fogging, reflective distortion and texture gradual change on a composite roof roadway drill hole wall image. Comprising the steps of hole wall image preprocessing and cylindrical surface expansion correction, near-far contrast difference calculation and equivalent driving quantity thermodynamic diagram generation, imaging quality and contour continuity weight construction, multi-source feature fusion depth segmentation and boundary contour extraction, contour and mask write-back weight self-updating and closed-loop iteration convergence control. According to the method, through near-far side comparison driving thermodynamic diagrams, imaging quality and contour continuity weight constraints and in combination with depth segmentation closed-loop iteration, the boundary contour of the potential structural plane of the separation layer is stably extracted, and false edges, fractures and horizon drifts caused by fogging water films and crack drill marks are reduced.

This invention discloses an image fusion method based on shifted window attention and semantic-driven dual adversarial approach. The method includes the following steps: Step 1, constructing an adaptive feature extraction network based on a dual-stream shifted window Transformer as a generator, utilizing a cross-window attention mechanism to achieve deep interaction and fusion of infrared image features and visible light image features; Step 2, constructing a multi-style dual discriminator decoupled from target and texture, introducing a target mask, and establishing brightness discrimination paths for salient infrared targets and texture gradient discrimination paths for visible light backgrounds; Step 3, introducing a semantic-driven meta-feature embedding feedback mechanism, utilizing a pre-trained target detection network to extract high-level semantic features, and constructing a semantic consistency loss to guide generator parameter updates; Step 4, constructing a joint objective function including content loss, dual adversarial loss, and semantic loss based on a mask weighting strategy, and performing end-to-end training on the network; Step 5, inputting the image to be fused into the trained generator, and outputting the fused image. This invention solves the problems of limited feature extraction and modal information conflict in traditional visible light and infrared fusion methods, significantly improving the detection accuracy of the fused image in machine vision tasks while balancing infrared high-brightness targets and clear background textures.

The application discloses a kind of based on edge-boundary conflict real scene three-dimensional model quality evaluation method, system, terminal and medium, method includes: obtaining real scene three-dimensional model, the gradient weighted normalization edge conflict index of each edge in real scene three-dimensional model is determined by texture gradient analysis, edge conflict index is used to reflect the degree that the geometric edge of real scene three-dimensional model is across domain semantic boundary in texture image space;Based on the edge conflict index of gradient weighted normalization, the face level edge conflict index of each triangular facet in real scene three-dimensional model is determined, face level edge conflict index is used to reflect the overall quality of triangular facet of real scene three-dimensional model;Establish grid level statistical evaluation framework, based on face level edge conflict index and grid level statistical evaluation framework, output quality evaluation result.The application is realized by constructing multilevel quality evaluation system, and overall quality characterization and low quality facet automatic identification from edge level, facet level to overall grid level of real scene three-dimensional model.

The invention belongs to the technical field of image analysis, and particularly relates to a waste water circulation treatment monitoring method and system for a sand separator, and the method comprises the steps: collecting a surface image of a waste water pool, enhancing the texture difference between foam and waste water through the technologies of local variance calculation, self-adaptive histogram equalization and the like, and calculating the surface image of the waste water pool; accurately segmenting a foam area by using an active contour model based on an enhanced texture gradient map; the foam coverage rate is obtained by calculating the area surrounded by the contours after convergence, the average brightness and the average texture of a foam area are extracted, and a foam thickness index is obtained through weighted fusion; finally, the system executes a hierarchical cooperative control strategy according to the two key indexes of the foam coverage rate and the thickness index, the defoaming device is intelligently started, stopped or adjusted, and automatic, refined and energy-saving management of foam is achieved.

The present application relates to the technical field of optical detection, and particularly relates to a multi-light-source fusion optical complex workpiece defect detection system and method; the method comprises the following steps: calculating a three-dimensional depth gradient field for a three-dimensional depth map, calculating a two-dimensional texture gradient field for a two-dimensional bright field image, and performing pixel-level spatial registration on the three-dimensional depth gradient field and the two-dimensional texture gradient field; inputting the registered three-dimensional depth gradient and two-dimensional texture gradient into a decision-level fusion model, calculating a comprehensive confidence degree of each pixel belonging to a defect by using weighted voting or DS evidence theory, and generating a defect probability map; performing threshold segmentation and connected domain analysis on the defect probability map, and outputting defect position, category and size information; by inputting the pixel-level spatially registered three-dimensional depth gradient and two-dimensional texture gradient into the decision-level fusion model, the problems of large registration error of existing multi-modal data and difficulty in distinguishing defect categories are solved.

This invention relates to the field of watershed topographic monitoring technology, providing a method and system for intelligent watershed monitoring based on UAV oblique photography. The method includes the following steps: Based on a watershed digital elevation model, the flight altitude is dynamically adjusted according to topographic undulations to acquire oblique photographic images; potential feature points are initially screened on each image, and the quality index of each potential feature point is determined based on neighborhood conditions and elevation information; the final retained feature points are determined based on a quadtree algorithm, topographic conditions, and the quality index; the texture information of the surrounding images of the final retained feature points is encoded, and this encoded information is used for comparison and matching in different images; feature point matching is performed, and the matching confidence of each feature point is determined. When the matching confidence is lower than a confidence threshold, dynamic accuracy compensation is performed based on structure tensor and texture gradient to determine the 3D model. This invention can promptly detect and correct errors that occur during the matching process, improving the accuracy and quality of the 3D model.

The invention relates to the technical field of computer vision, in particular to an infrared zoom lens control method and system based on machine vision, and the method comprises the following steps: carrying out the multi-scale sampling of an infrared image, extracting a thermal texture gradient, and generating a multi-scale thermal texture gradient amplitude matrix; calculating a cross-scale co-location ratio to obtain a cross-scale thermal texture coupling rate matrix, carrying out differential detection extreme value migration to recognize a dominant scale and generate a focal length regulation and control weighting parameter, carrying out weighted fusion on multi-scale energy and mapping a focal plane compensation stepping amount, and carrying out dynamic compression and fine tuning stepping to generate a lens focal length adjustment instruction. According to the method, direction judgment is completed through multi-scale texture gradient construction and cross-scale coupling extraction, dominant scale weight is determined in combination with extreme value migration, energy fusion and integral extraction are completed through regulation and control based on global gradient evolution, compression compensation stepping is generated under noise suppression, and continuous output and smooth convergence are achieved. And the jumping and lagging problems in the zooming process are improved.

The application discloses a spring assembly multi-view visual feature reconstruction method for complex working conditions, and particularly relates to the fields of computer vision and industrial detection; multi-view images of a target region are acquired, and a candidate spring region is extracted; main shaft direction analysis is carried out based on local texture gradient features, and a three-dimensional sparse point cloud is constructed; a physical structure constraint model is introduced to perform morphology correction on the point cloud, and an intermediate feature model is generated; high-dimensional geometric features are extracted through a multi-scale visual attention network, and a fine three-dimensional point cloud model containing microstructure differences is reconstructed; candidate component confidence is sorted according to geometric consistency evaluation results, pseudo-identified objects are removed, and a final spring set is output; the application can realize high-precision structure recovery under the conditions of occlusion, stacking and light interference, and has strong industrial adaptability.

The invention discloses a plastic production quality defect detection method based on image detection, and relates to the technical field of image detection, and the method comprises the following steps: executing refractive index estimation and light path inversion calculation on light offset characteristics, and generating a light path correction image; performing illumination normalization, local contrast enhancement and texture gradient extraction on the light path correction image to form a candidate area image containing a defect candidate area; and inputting the candidate region image into a self-supervised reversible reconstruction network for secondary reconstruction to generate a corresponding candidate region reconstruction image, and calculating a defect reconstruction difficulty score based on a pixel residual between the candidate region image and the candidate region reconstruction image. According to the method, the candidate region image is input into the self-supervised reversible reconstruction network, and the defect reconstruction difficulty score is constructed based on the reconstruction error, so that the refined quantitative expression of the authenticity performance of the candidate region in the image detection process is realized.

The application relates to the technical field of image compression transmission, and discloses a satellite image transmission monitoring station image compression transmission method and system for a reservoir, which comprises the following steps: obtaining an original image, removing system errors to obtain a clear image; extracting a waveband edge intensity and a texture gradient, performing a connected domain morphological close operation after double-threshold screening, and determining a core target region boundary; expanding an adjacent pixel outward by counting a boundary pixel histogram to obtain an expansion region; calculating a saliency map based on the region, screening high saliency pixels, performing connected component marking, and generating a protection area mask; performing wavelet transformation on the expansion region, encoding high-fidelity data according to a mask low quantization step, high quantization encoding of a background region to generate compressed data, and fusing the high-fidelity data and the compressed data to obtain a hybrid compressed image; after verification, when a receiving end decodes, core data is losslessly restored, a background region is reconstructed in detail, and a complete image is output. The method can solve the problem of image distortion.

The application relates to the technical field of digital image processing, and discloses a terminal software-oriented image compression and restoration method and device, which comprises the following steps: acquiring an image to be processed and a window interaction state sequence of current terminal software; extracting local texture gradient features of the image to be processed; and extracting corresponding low-frequency basic residual errors and high-frequency directional residual errors; asymmetrically encoding the low-frequency basic residual errors and the high-frequency directional residual errors according to a frequency domain quantization step to generate a structured compression code stream; analyzing the structured compression code stream to acquire a basic reconstruction image and local detail compensation features; dynamically adjusting fusion weights of the local detail compensation features according to a real-time rendering frame rate of the current terminal software, and superimposing the local detail compensation features to the basic reconstruction image to generate a target restored image; and the application can improve the balance between high-resolution visual experience and limited terminal computing resources.

The invention discloses a spring assembly multi-view visual feature reconstruction method oriented to complex working conditions, and particularly relates to the field of computer vision and industrial detection. Obtaining a multi-view image of the target area, and extracting a candidate spring area; performing principal axis direction analysis based on local texture gradient features, and constructing a three-dimensional sparse point cloud; introducing a physical structure constraint model to perform morphological correction on the point cloud to generate an intermediate feature model; high-dimensional geometric features are extracted through a multi-scale visual attention network, and a fine three-dimensional point cloud model containing tiny structure differences is reconstructed; sorting the confidence coefficients of the candidate components according to a geometric consistency evaluation result, rejecting a false identification object, and outputting a final spring set; according to the method, high-precision structure recovery can be realized under the conditions of shielding, stacking and illumination interference, and the method has relatively high industrial adaptability.

The invention provides a cultivated land boundary identification method and system, and the method comprises the steps: collecting a remote sensing image of a target cultivated land, carrying out the cultivated land semantic segmentation of the remote sensing image, obtaining an initial cultivated land probability graph, and determining a color-texture gradient magnitude graph of the remote sensing image; the category probability gradient of each pixel in multiple directions in the cultivated land probability graph and the color-texture gradient magnitude graph of the remote sensing image are initialized, and a boundary significance graph is constructed; according to the semantic probability gradient magnitude and the color-texture gradient magnitude of each pixel point in the boundary saliency map, determining the semantic-geometric feature mismatch degree of each pixel point, and performing adaptive filtering on the boundary saliency map through the semantic-geometric feature mismatch degrees of all the pixel points; and generating a cultivated land boundary vector of the target cultivated land based on the filtered boundary saliency map. By adopting the scheme of the invention, the dislocation of the semantic boundary and the visual edge in the cultivated land boundary image recognition process can be quantified.

The invention relates to the technical field of industrial machine vision detection, and discloses a dispensing quality detection method and system of a dispensing detection machine. The method comprises the following steps: acquiring a multispectral image of a dispensing area, and fusing based on colloid transmission characteristics to obtain a multiband image set; performing median filtering and illumination normalization on the multiband image set to obtain a corrected image set; performing principal component analysis on the corrected image set to extract a band difference vector, and inputting a texture gradient into a convolutional neural network to obtain a fusion feature map; performing threshold comparison and morphological screening on the fused feature map to obtain a preliminary defect mark; intercepting candidate image blocks, extracting reflectivity and texture features, and determining final defect information by using a support vector machine; and performing closed-loop adjustment on the preprocessing parameters based on the detection confidence coefficient variance. According to the invention, the interference of complex illumination and variable material environment can be overcome, and high-precision and high-stability detection of tiny dispensing defects can be realized.

The invention relates to the technical field of image recognition, in particular to an automobile stamping component quality detection method based on image recognition, and the method comprises the steps: enabling a brightness variance disturbance sequence and a texture gradient change sequence to present statistical distribution characteristics through the weight distribution construction of a Bayesian neural network; a probability interval of a defect area is formed by multiple outputs generated by multiple random forward calculations, so that the area has quantitative dispersion expression in a fluctuation environment, and fluctuation risks of cracks, indentations and wrinkles are presented in an interval form; the introduction of the convolutional neural network in a region symbolization process enables the area, curvature and texture balance degree of a region block to be presented in a hierarchical expression mode in a multi-channel response graph, and a convolution kernel generates different response amplitudes for region intensity, texture direction and boundary bending during local sliding. And enabling the same structure to form a response combination containing local and overall differences at the same time under multiple scales, so that the successive hierarchy during defect multi-term combination is presented in a symbol chain sequence mode.

The present application belongs to the technical field of image analysis, and particularly relates to a wastewater recycling treatment monitoring method and system for a sandstone separator, which comprises the following steps: collecting a surface image of a wastewater pool, using local variance calculation and adaptive histogram equalization and other technologies to enhance the texture difference between foam and wastewater, and then using an active contour model based on an enhanced texture gradient image to accurately segment the foam area; calculating the area surrounded by the converged contour line to obtain the foam coverage, and extracting the average brightness and average texture of the foam area to be weighted and fused into a foam thickness index; finally, the system executes a hierarchical collaborative control strategy according to the two key indicators of the foam coverage and the thickness index, intelligently starts and stops or adjusts the defoaming device, and realizes the automatic, refined and energy-saving management of the foam.

The invention relates to the field of industrial detection, and discloses a plastic injection molding defect real-time detection method and system and a storage medium. The method comprises the following steps: synchronously imaging a defect area of a plastic injection molding part through a multi-angle camera array, and pre-processing to generate an enhanced defect candidate point feature set; carrying out three-dimensional reconstruction through a stereoscopic vision algorithm and combining with convolutional neural network classification to generate a defect type set; curvature analysis and surface texture gradient analysis are carried out on the defect type set, and shadow texture analysis and texture roughness calculation results are fused to generate a defect cause parameter identification set; adopting a dynamic tracking algorithm to generate a defect space position evolution trend information set; and fusing the defect type set to optimize injection molding process parameters to obtain a mold design improvement data set. Through fusion of multiple technologies such as multi-angle imaging, stereoscopic vision, the convolutional neural network and dynamic tracking, the problems of low detection precision, weak interference resistance and no closed-loop optimization in the prior art are solved, and the injection molding defect detection efficiency and the process improvement scientificity are improved.

This invention discloses a dynamic face texture generation method based on facial expression semantic features, belonging to the interdisciplinary field of computer vision and graphics. It includes: extracting geometric parameters from a parametric face model tracking video, driving model deformation, and anchoring 3D Gaussian points; constructing a dynamic texture generation network, using an identity latent code as a basis, and modulating the texture feature map layer by layer using facial expression parameters through a feature modulation module to output a dynamic UV texture map; calculating the gradient magnitude of the UV texture map, and adaptively encrypting the Gaussian points based on the texture gradient and geometric gradient; obtaining Gaussian point attributes based on UV coordinate sampling, and inputting them into a differentiable rasterizer renderer after viewpoint-related color compensation to generate an image. This achieves dynamic texture generation that is linked to facial expressions, significantly improving the dynamic realism and detail fidelity of digital human reconstruction, and solving the problem of high-frequency dynamic detail loss caused by static texture representation in existing monocular video 3D face reconstruction.

The present application relates to machine vision and image processing technical field, specifically to a kind of workpiece defect feature point's visual image enhancement processing method, including original gradient field generation step: the information of workpiece surface illumination reflection is converted into two-dimensional vector field;Local texture extraction step: introduce structure tensor field and obtain local texture direction field by eigenvalue decomposition;Background flow field construction step: carry out smoothing regularization processing to generate background flow field removed local disturbance;Signal geometry separation step: original gradient field is projected and decomposed into along-texture and inverse-texture gradient component;Gain reconstruction enhancement step: component is handled and reconstructed using differential gain control, and output enhanced image;The present application utilizes the adaptive characteristics of structure tensor field, and realizes the accurate stripping of weak defect signal under strong texture interference.

The invention discloses an exposure control method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring a current frame image; extracting multi-scale texture gradient features of the current frame image; constructing a distance sensing model based on camera internal parameters, and mapping the multi-scale texture gradient features into spatial distance weights according to the distance sensing model; generating a significance map according to the spatial distance weight; determining exposure control parameters according to the saliency map; and adjusting image exposure according to the exposure control parameters. By analyzing texture gradient features of an image and utilizing internal reference of a camera to perform distance perception mapping, exposure of an area closer to the camera and more likely to be a main body in a picture can be actively deduced and preferentially optimized under the condition that targets such as a human face are not successfully detected; the problem of deadlock of the whole exposure process caused by failure of initial detection is avoided, and the fundamental problem that exposure control cannot be started under severe illumination conditions such as backlight is solved.

The invention relates to the technical field of chip image analysis, and discloses an intelligent chip defect detection method and system, and the method comprises the steps: obtaining a wafer image and process metadata; constructing a process gradient field model representing a global process drift trend, and dividing the wafer into equivalent process regions; quickly screening and identifying suspicious regions based on an equivalent process region generation template; constructing a reference candidate set, dynamically generating a local adaptive reference for the suspicious region, and obtaining a fine residual image; and performing multi-level feature extraction and classification based on the fine residual error, and outputting defect information. The system comprises a data acquisition module, a modeling processing module, a rapid screening module, a fine diagnosis module and a defect discrimination module. According to the method, the wafer process gradient field model is constructed, chip grain texture gradual change is adapted, drift misinformation is effectively overcome, high-throughput full inspection is guaranteed, and meanwhile accurate identification of low-contrast macroscopic and tiny defects of the chip is achieved.

The invention discloses a cerebral hemorrhage intelligent aid decision-making system based on knowledge graph and multi-modal fusion, and particularly relates to the technical field of medical big data, comprising a hematoma texture feature extraction module, a graph edge vector quantization module, an indication fluctuation sequence calculation module, a graph decision joint calibration module and a diagnosis and treatment risk quantitative evaluation module. According to the method, real-time difference fluctuation of intracranial pressure and systolic pressure is utilized, the depth of a decision tree is dynamically expanded, the map path weight is corrected, a self-adaptive reasoning mechanism based on the physical sign time-varying characteristic is constructed, the Euclidean distance between the decision tree and map nodes is calculated, radius judgment is executed, the spatial consistency of logic deduction and knowledge retrieval is locked, and the reliability of the decision tree is improved. And ensuring that the risk assessment model is synchronously calibrated along with the physiological state of the patient and the texture gradient.

The application relates to the technical field of industrial machine vision detection, and discloses a dispensing quality detection method and system of a dispensing detection machine. The method comprises the following steps: collecting a multi-spectrum image of a dispensing area, fusing a multi-band image set based on the transmission characteristics of a glue body; performing median filtering and illumination normalization on the multi-band image set to obtain a corrected image set; performing principal component analysis on the corrected image set to extract a band difference vector, combining a texture gradient to input a convolutional neural network to obtain a fused feature map; performing threshold comparison and morphological screening on the fused feature map to obtain preliminary defect labels; intercepting a candidate image block and extracting reflectivity and texture features, determining final defect information by using a support vector machine; and adjusting preprocessing parameters based on detection confidence variance in a closed loop. The application can overcome the interference of complex illumination and variable material environments, and realize high-precision and high-stability detection of tiny dispensing defects.

The present application relates to the technical field of medical digital imaging, and discloses a CT scanning parameter self-adaptive control and adjustment system based on texture decomposition reconstruction, which comprises a sliding window data acquisition module, a fast texture state sensing module, a dual variable saturation analysis module and a dual-channel dynamic impedance matching control module. The system uses a sliding window to real-time intercept local projection data, extracts a second dual variable representing texture gradient through an original-dual iterative algorithm, and calculates the saturation of the second dual variable within a constraint boundary to generate a feedback error. The control module responds to the error to perform dual-channel synchronous adjustment: the first channel adjusts the tube current of an X-ray generating device, and the second channel dynamically adjusts the regularization parameter of a reconstruction model according to the tube current value. The present application realizes dynamic impedance matching of physical photon flux and algorithm regularization intensity, ensures the balance of image texture and noise under different attenuation regions, and reduces the radiation dose while ensuring the diagnostic image quality.