22 results about "Edge density" patented technology

The invention relates to the field of defect detection, in particular to a wallpaper defect detection method and system based on machine vision. The method comprises the following steps: analyzing acquired to-be-detected wallpaper image data to obtain regional texture variance, edge density and brightness gradient indexes, and generating a scale image layer set; extracting salient edge points in each scale image and generating an edge anchor point set; analyzing the scale image layer set and the edge anchor point set to obtain a texture difference index, and generating a texture feature map set according to the edge anchor point set; processing the texture feature map set to generate a texture significance distribution map; based on the texture significance distribution diagram, main direction distribution is extracted, a direction residual error model is constructed, and a residual error diagram and a direction difference distribution diagram are generated; and establishing a region scoring matrix, and generating a wallpaper defect credibility distribution map and a defect classification image output set on the basis of the region scoring matrix. The wallpaper defect detection precision can be improved.

The invention discloses a multi-mode resume layout self-adaptive analysis method and system based on a large model, and particularly relates to the technical field of self-adaptive analys.The multi-mode resume layout self-adaptive analysis method comprises the steps that firstly, resume files of multiple formats uploaded by a user are converted into standard image formats in a unified mode, and edge density calculation and language recognition are conducted on images; and judging whether to perform image enhancement processing or not according to the calculated image enhancement coefficient. Then, text information is extracted through the OCR technology, multi-modal representation is constructed through fusion of a visual feature extraction model and a large language model, and paragraph positions are corrected in combination with a diffusion layout model; the system further adopts a multi-label classification model to identify a functional region, constructs a similar sample set by matching historical resumes, calculates the weight of each region based on HR attention and content deviation, performs sorting and length optimization on resume contents, and finally generates a personalized resume structure which is clear in structure and prominent in expression. The system is suitable for resume analysis and content recommendation in a complex layout and multi-language environment.

The invention discloses a mobile terminal real-time ray tracing noise reduction method and system based on edge perception, and the method comprises the steps: carrying out the multi-channel edge detection of the G-buffer information of an input frame, and extracting an image edge; generating a complete edge density map of the current extraction frame; dynamically dividing the pixels into calculation groups with different kernel sizes according to the edge density value; performing pixel clustering in each kernel size calculation group by using local depth gradient information; for each sub-cluster, brightness (obtained by albedo conversion), surface normal and depth information in each pixel G-buffer in the sub-cluster are obtained, a filter weight corresponding to each sub-cluster is obtained by using a lightweight multilayer perceptron network, and a global weight lookup table is generated; and quantizing the continuous feature space of the frame to be detected into a discrete index, realizing weight retrieval by using the global weight lookup table, and finally executing spatial filtering according to the allocated kernel size and weight to output a noise-reduced image. The real-time performance and the quality are both considered, and the noise reduction efficiency is improved on the premise that the noise reduction effect is guaranteed.

The invention discloses a shooting method and wearable equipment. The shooting method is applied to the wearable equipment with a camera. The method comprises the steps that a preview frame corresponding to a shooting target is obtained, and the preview frame is single-frame image data in a target preview flow collected by a camera and is not displayed in the collection process; fusing the preset visual features of the preview frame to obtain a decision score of the preview frame; the preset visual features comprise at least two items of a target length-width ratio, an edge density ratio and an image scene category; and analyzing the decision score based on a dynamic threshold strategy, and adjusting the shooting direction of the shooting target. In a scene that a user cannot preview a shot picture, multi-dimensional visual feature analysis is carried out on a preview frame in a real-time preview stream shot by the wearable device, after visual features of all dimensions are fused, a decision score is obtained, then the decision score is analyzed through a dynamic threshold strategy, and in a scene that the shot picture cannot be previewed, the preview frame in the real-time preview stream shot by the wearable device is subjected to multi-dimensional visual feature analysis. And the shooting direction of the shooting target is adjusted.

The application discloses a multi-source data feature extraction and fusion method and system for power grid infrastructure construction, and relates to the technical field of electronic data processing.The method comprises the following steps: obtaining monitoring images and noise data of a construction site and performing pretreatment; extracting color, texture and edge density features from the images, generating an initial segmentation mask of an engineering area through weighted fusion; performing morphological operation, area filtering and solid degree filtering optimization on the mask, extracting an engineering area area proportion progress index and shape features based on the final mask; extracting frequency band energy, power spectrum density peak value and statistical features from the noise data; and finally, fusing the image features and noise features through an adaptive weighted fusion algorithm based on variance and correlation coefficient to generate a final feature vector.The application improves the accuracy and reliability of state perception through multi-source data processing and feature-level fusion, and provides a high-quality data basis for subsequent intelligent analysis.

The invention discloses a multi-source data feature extraction and fusion method and system for power grid infrastructure construction, and relates to the technical field of electronic data processing, and the method comprises the following steps: obtaining a monitoring image and noise data of a construction site, and carrying out the preprocessing of the monitoring image and noise data; extracting color, texture and edge density features from the image, and generating an engineering region initial segmentation mask through weighted fusion; performing morphological operation, area filtering and solid degree filtering optimization on the mask, and extracting an engineering region area proportion progress index and a shape feature based on the final mask; extracting frequency band energy, a power spectral density peak value and statistical characteristics from the noise data; and finally, through an adaptive weighted fusion algorithm based on variance and correlation coefficients, fusing the image features and the noise features to generate a final feature vector. Through multi-source data processing and feature level fusion, the accuracy and reliability of state perception are improved, and a high-quality data basis is provided for subsequent intelligent analysis.

The invention is suitable for the technical field of flange detection, and provides a flange and an end surface flatness detection method thereof, and the method comprises the steps: S1, surrounding light source irradiation: irradiating the end surface of the flange through at least one annular light source, and enabling the end surface of the flange to present uniform reflected light; s2, collecting a reflection image of the flange end face through an industrial camera; s3, edge detection analysis: processing the reflection image by using a Canny edge detection algorithm or a Laplace operator, and detecting gray level change and edge mutation on the flange end face; judging whether tiny scratches, local sudden changes or tiny unevenness exist on the flange end face or not by analyzing the area with the gray value changing drastically in the image; s4, comprehensive defect evaluation: according to the results of S2 and S3, if the variable coefficient CV is greater than a threshold value T1, or any one of edge indexes (including edge density and gray level change intensity E) is greater than a corresponding threshold value T2 or T3, determining that the flange end face has defects, and determining that the detection result is unqualified; otherwise, judging that the product is qualified.

The application discloses an analysis method for a borehole image under low turbidity, comprising the following steps: collecting a three-channel hole wall image by using an industrial camera installed at the front end of a detection probe and synchronously acquiring a hole depth position, fusing to generate a gray-scale image, performing edge detection on the gray-scale image to obtain a binary edge image, performing texture suppression on the binary edge image by using a filter to obtain an edge image after texture suppression, calculating edge density by using edge pixels in the edge image and generating an edge density heat map, acquiring a particle coverage area rate and a particle size distribution factor based on the heat map, further calculating an equivalent turbidity index, dividing a turbidity level according to a reference interval of the ETI and identifying whether hole washing is qualified, and realizing rapid quantitative evaluation of particle residue on a hole wall caused by incomplete hole washing by means of multispectral fusion imaging and edge density analysis, effectively suppressing interference of original textures on the hole wall, providing a repeatable and comparable quantitative index for hole washing quality, supporting on-site real-time decision on whether to continue hole washing, and significantly saving construction time.

The invention relates to the technical field of image processing, can be applied to business system platforms of financial science and technology, medical health and the like, and discloses a target sample identification method, device and equipment and a computer readable storage medium, the method comprises the following steps: generating a global feature vector according to scene category information and space structure information in each sample image; calculating the similarity between the global feature vectors of the target image and the other images, selecting the images with the similarity greater than a first threshold value as similar images, and converging all the similar images to generate a first candidate image group; calculating a background score according to the edge density feature and the texture feature in the background region of each sample image, and deleting the images with the background scores smaller than a second threshold value to form a second candidate image group; and selecting a sample image pair of which the number of matched key points is less than a third threshold and the similarity is greater than a fourth threshold from the second candidate image group as a target sample. According to the invention, the accuracy of identifying the target sample is improved.

An image content analysis method is applied to an image analysis apparatus and includes acquiring an image, utilizing an edge detection technology to compute an edge density of the image, utilizing a texture detection technology to compute a texture richness of the image, and analyzing the edge density and the texture richness to generate a richness score of the image.

The invention discloses an image storage method, image storage equipment and a computer storage medium, and relates to the technical field of image processing and storage. The method comprises the following steps: acquiring original data through an image acquisition sensor, calculating local contrast and edge density, performing fusion to generate a visual saliency map, and performing region segmentation to obtain an average visual saliency value of each region; calculating a statistical variance and a global peak value according to the values of all the areas, and generating an adaptive rate distortion reference value; comparing each region value with a reference value to determine a quantization step size and a visual saliency level, and performing adaptive coding on the region to form a composite data stream; calculating the proportion of the geometric center to the area of each region, packaging the proportion into metadata, and packaging and storing the metadata and the coding stream; and constructing a mapping index from the region identifier to the physical address and a reverse index based on the visual saliency level. The method can improve the utilization rate of the storage space and optimize the image quality of the important region.

The invention discloses a scene perception and multi-feature fusion-based adaptive white balance method, which comprises the following steps of: inputting an RGB (Red, Green and Blue) image to carry out multi-feature calculation, and screening out effective image blocks meeting the conditions of a saturation range, a brightness range, a color variance threshold value and an edge density threshold value through a dynamic threshold value; if the number of the effective blocks is lower than a preset threshold value, starting a parameter dynamic adjustment mechanism to carry out secondary block screening; if the number of effective blocks after secondary block screening is still insufficient, activating a standby strategy to perform multi-feature fusion correction, and outputting a white balance image; after the effective blocks are sufficient, switching to a main strategy for processing, and outputting a white balance image through gain calculation and constraint; according to the scheme, the problem of color cast when the image is lack of a white / gray region is solved, automatic parameter adjustment under the condition that the illumination condition changes drastically is achieved, and noise caused by gain amplification can be restrained in a low-illumination scene.

The application discloses a bridge crack detection method based on image recognition and concretely relates to the field of bridge detection, which is used to solve the problems of misjudgment, missed detection and crack breakage caused by the similar shape of the code or steel stamp characters on the bridge surface and the crack fine lines; the edge graph is generated from the original image, the character candidate area is limited based on the edge density and the contrast on both sides of the edge, the character mask and the character skeleton are constructed in the character candidate area, the cost graph is constructed by combining the multi-directional line top-hat transformation inside the character mask, the minimum cost path is quantified to pass through the uniform cost and attach the uniform distance, the classification model is used to output the crossing coefficient to separate the crossing candidate set and the homologous candidate set, the sample block matching is used to generate the background graph without characters and the residual fine line graph in the character repair area to complete the missing section of the crossing candidate set and remove the corresponding pixels of the homologous candidate set, and the crack result pixel graph is output.

This invention belongs to the field of image data processing technology, specifically relating to a multi-feature adaptive cloud detection method based on satellite cloud images. The method includes: acquiring and preprocessing satellite cloud images to obtain grayscale images; extracting image feature vectors from the grayscale images, wherein the image feature vectors include contrast, edge density, texture complexity, mean brightness, and mean gradient magnitude; matching detection methods based on contrast features, edge density features, and texture complexity features in the image feature vectors; generating cloud masks based on the detection methods, wherein when K-means clustering is used as the detection method, the brightest cluster is selected to generate the cloud mask; when an adaptive threshold segmentation method is used, candidate masks are generated based on a first threshold and a second threshold formed by the standard deviation and mean of the pixel values ​​of the grayscale image, and the optimal candidate mask is selected as the optimal mask; and using morphological opening and closing operations to remove isolated noise points and fill holes in the optimal mask to obtain the final mask.

The invention belongs to the technical field of image data processing, and particularly relates to a satellite cloud picture-based multi-feature adaptive cloud detection method, which comprises the following steps of: acquiring and preprocessing a satellite cloud picture to obtain a grayscale image; extracting an image feature vector of the grayscale image, wherein the image feature vector comprises a contrast ratio, an edge density, a texture complexity, a brightness mean value and a gradient magnitude mean value; matching a detection method based on a contrast feature, an edge density feature and a texture complexity feature in the image feature vector; a cloud mask is generated based on a detection method, and when a K-means clustering method is adopted as the detection method, the brightest cluster is selected to generate the cloud mask; when the adaptive threshold segmentation method is adopted, candidate masks are generated based on a first threshold and a second threshold formed by a standard deviation and a mean value of pixel values of the grayscale image, and an optimal mask is selected from the candidate masks as an optimal mask; and removing isolated noisy points and filling holes for the optimal mask by adopting morphological opening and closing operation to obtain a final mask.

The application discloses an image storage method, an image storage device and a computer storage medium, and relates to the technical field of image processing and storage. The method comprises the following steps: acquiring original data through an image acquisition sensor, calculating local contrast and edge density, and fusing to generate a visual saliency map; obtaining an average visual saliency value of each region through region segmentation; calculating statistical variance and a global peak value according to the value of all regions to generate an adaptive rate distortion reference value; comparing the value of each region with the reference value to determine a quantization step and a visual saliency level, adaptively encoding the region to form a composite data stream; calculating the geometric center and area proportion of each region, packaging the data into metadata, and packing and storing the metadata with the encoded stream; and constructing a mapping index of region identification to a physical address and an inverted index based on the visual saliency level. The application can improve the storage space utilization rate and optimize the image quality of important regions.

The invention relates to the field of high-resolution remote sensing image analysis based on artificial intelligence, and discloses a high-resolution satellite data semantic analysis and intelligent identification system and method, and the method comprises the steps: carrying out the texture complexity, edge density and significance response analysis of a high-resolution satellite image; generating a multi-scale adaptive grid unit and establishing a spatial index; loading a corresponding multi-temporal image, and generating a multi-scale spatial-temporal feature sequence; determining a neighborhood grid set according to the multi-scale spatio-temporal feature sequence and the spatial index, generating a spatio-temporal incidence matrix and predicting a semantic change trend of a main grid, performing scale alignment and local disturbance suppression on multi-temporal features, and constructing a cross-spatio-temporal stable reference framework; fusing the cross-spatio-temporal stable reference framework, the semantic change trend, the spatio-temporal incidence matrix and the current frame preliminary semantic judgment to form a comprehensive confidence coefficient matrix; and adjusting the key feature weight and the matching parameter, and outputting a corrected grid semantic recognition result. The method has the advantage of improving the recognition precision.

The invention relates to a colorectum navigation correction method based on intra-operative image fusion, which comprises the following steps of: judging an image continuity state by analyzing tissue edge continuity and a gray level change trend between intra-operative image frames, if a breakpoint is detected, starting an image backtracking mechanism, and selecting a stable frame sequence as fusion input; dividing the intra-operative image into a plurality of space blocks, calculating the structural complexity and edge density index of each block, and only reserving a credible region as a fusion candidate region; performing spatial arrangement logic analysis on anatomical structure labels identified in a fusion result, comparing the anatomical structure labels with a preoperative standard anatomical model, and if structural sequence abnormity or logic conflict exists, judging that the fusion registration is invalid and stopping navigation updating; after the fusion error is identified, analyzing the change of the main axis direction of the structure in the operation, deducing the rotation vector of the navigation offset, carrying out local self-rotation adjustment on the image in the operation, re-aligning the pre-operation planning coordinate system, correcting the navigation position information, and using the corrected result for updating the navigation system.

The present application relates to the field of defect detection, in particular to a wallpaper defect detection method and system based on machine vision; the method comprises the following steps: analyzing the obtained wallpaper image data to be detected, obtaining regional texture variance, edge density and brightness gradient index, and generating a scale image layer set; extracting significant edge points in each scale image and generating an edge anchor point set; analyzing the scale image layer set and the edge anchor point set to obtain a texture difference index, generating a texture feature map set based on the edge anchor point set; processing the texture feature map set to generate a texture saliency distribution map; based on the texture saliency distribution map, extracting a main direction distribution, constructing a direction residual model, generating a residual map and a direction difference distribution map; establishing a regional score matrix, and generating a wallpaper defect credibility distribution map and a defect classification image output set based on the regional score matrix. The present application can improve the detection accuracy of wallpaper defects.

This invention discloses a method for detecting small targets in remote sensing images against complex backgrounds, comprising the following steps: Step 1: Acquiring the original remote sensing image data and corresponding metadata of the area to be detected; Step 2: Performing grayscale distribution stretching, band normalization, and local brightness balancing processing; Step 3: Calculating local texture repetition, edge density, regional entropy, high-frequency energy in the frequency domain, and directional gradient; Step 4: Constructing a background complexity constraint map; Step 5: Extracting the low-level feature maps of standardized remote sensing sub-patterns and performing matrix operations with the background complexity constraint map; Step 6: Inputting the initial feature tensor into an improved CrossFormer model to generate a cross-scale small target enhancement representation sequence; Step 7: Mapping the detection results to the original remote sensing image coordinate system to obtain the final detection result. This invention achieves accurate detection of small remote sensing targets through the improved CrossFormer model.

The invention discloses an insulating cylinder defect detection method and a storage medium. The method comprises the following steps: acquiring multi-angle image data of an insulating cylinder, and respectively preprocessing edge and center features through anisotropic diffusion filtering and a CLAHE algorithm to generate a high-contrast image; enhancing edge detection by adopting an improved Scharr operator, adaptively dividing rectangular sub-blocks based on an edge density function, and ensuring that each sub-block comprises at least one potential defect unit; extracting multi-scale LBP entropy value features by using wavelet transform, and constructing a support vector machine classification model to realize accurate defect classification; and finally marking the position and the size of the defect through a visualization algorithm. According to the method, a dynamic partitioning strategy and multi-scale texture analysis are fused, and the technical problems that a traditional method is high in false detection rate and small defects are missed are solved.

The invention discloses a power distribution network defect image definition screening method and system based on multi-feature fusion, and belongs to the interdisciplinary field of power system intelligent detection and computer vision technology, and the method comprises the following steps: 1, inputting an RGB image collected by a power distribution network unmanned plane in an inspection manner; step 2, basic definition feature extraction is carried out through three types of airspace features of Laplace variance, Tenenggrad gradient and Sobel gradient energy; 3, calculating the average brightness of the image, and eliminating the interference of illumination fluctuation on definition evaluation; 4, obtaining a binary edge image of the image through Canny edge detection, and calculating an edge density and an edge gradient mean value; 5, after normalization processing is carried out on the three types of basic features, a weighted model is constructed in combination with an illumination compensation factor and an edge confidence coefficient weight; and step 6, setting a definition threshold value, and if the comprehensive score reaches or exceeds the threshold value, determining that the image is a clear image and keeping the image for subsequent defect identification.