27 results about "Gray level histogram" patented technology

The invention provides a coal impact tendency evaluation method which comprises the following steps: sampling a coal rock material on an engineering site, and processing the sample into a coal rock test piece with a preset size and a preset shape; carrying out a uniaxial compression experiment on the coal rock test piece, and recording a motion image of an ejection body on the coal rock test piece in the loading process of the coal rock test piece; comparing cosine similarities among the plurality of images in the moving images, and determining the image with the maximum damage degree according to the cosine similarities; analyzing gray level distribution of the image with the maximum damage degree to obtain a gray level histogram; identifying particle information of the broken particles in the image with the maximum damage degree; representing the damage degree of the coal rock material by using the particle information of the crushed particles; and evaluating the coal impact tendency by using the damage degree of the coal rock material. According to the evaluation method for the coal impact tendency disclosed by the invention, a new coal impact tendency evaluation index is provided, the research on the damage process of the coal rock material is perfected, and basic parameters are provided for subsequent early warning of damage of the coal rock material.

The invention discloses a tracking optimization method for detecting a low-speed small unmanned aerial vehicle target by a single-photon laser radar, belongs to the technical field of optical detection and target tracking, and aims to solve the problem of unstable target imaging and tracking caused by speckle noise interference in the prior art. The invention provides a speckle noise suppression method based on a vibration emission optical fiber and a space-time dynamic kernel density estimation algorithm, and the method is combined with an improved mean shift-Kalman filtering algorithm to achieve the tracking optimization of a low-speed small unmanned aerial vehicle target, and obtains an echo signal through a single-photon laser radar. Constructing a three-dimensional data matrix and reconstructing distance and intensity images through space-time filtering and kernel density estimation; and then, in combination with gray level histogram probability estimation, similarity measurement and Mean Shift iteration, target area tracking is realized, and finally, a target position is predicted and output by using Kalman filtering. The method is suitable for the fields of long-distance unmanned aerial vehicle detection and monitoring, low-altitude security defense early warning, civil airspace management and control, military anti-unmanned aerial vehicle systems and the like.

The invention provides a microscope automatic focusing method based on an end-to-end model, and relates to the technical field of optical instruments. The method comprises the following steps: shooting an image on a microscope glass slide through an image acquisition module, and carrying out image preprocessing through a gray histogram equalization algorithm; deploying an end-to-end neural network model at the cloud server, receiving the preprocessed image and predicting a first focusing position value; the first focusing position value is sent to an electric focusing module controlled by a stepping motor, and focusing is completed; and acquiring a second image, performing image preprocessing, predicting a second focusing position value by using the end-to-end neural network model, and sending a relative position between the second focusing position value and the first focusing position value to the electric focusing module to complete final automatic focusing. According to the method, the optimal focusing position is directly predicted from the microscopic image through the deep learning model, a traditional focusing scoring function or an image sequence analysis process is simplified, end-to-end efficient prediction is realized, and the focusing speed and precision are remarkably improved.

The invention relates to the technical field of liver image processing, in particular to an early screening model training method for non-alcoholic fatty liver. According to the method, the fat infiltration region is effectively screened out through the gray value of the pixel region and the region shape feature, and the fat layer saliency is obtained. And processing in the gray level histogram, analyzing the probability of each peak as a fibrosis peak except the highest peak, and finally judging the fibrosis peak and obtaining the fibrosis significance. Through the two features, automatic marking of the liver CT image is realized. And the early screening model of the non-alcoholic fatty liver can be trained through the automatically marked liver CT image. According to the embodiment of the invention, the features in the liver CT image are identified, through automatic marking, training of a non-alcoholic fatty liver early screening model is facilitated, and the training efficiency and the model accuracy are ensured.

The invention belongs to the technical field of image processing, and particularly relates to a boundary intrusion target detection method based on machine vision, and the method comprises the steps: obtaining a boundary monitoring video frame sequence, extracting a grayscale image, and constructing a background model image based on a historical frame sequence; according to the gray scale change of the pixel point in the time window, a time sequence fluctuation index is obtained through the standard deviation of the time sequence gray scale set and the overturning frequency of the gray scale difference value; according to texture distribution characteristics in a pixel point neighborhood, obtaining a spatial discrete entropy through a gray level histogram; obtaining a direction chaos index through the vector sum modulus and the algebraic modulus sum of the motion gradient vector; and correcting the original motion response diagram according to the time sequence fluctuation index, the spatial discrete entropy and the direction chaos index to obtain an intrusion confidence coefficient, and carrying out connected region detection according to the intrusion confidence coefficient. According to the method, the environmental dynamic interference is effectively inhibited by fusing the time sequence oscillation, the space texture and the motion direction, and the accuracy of boundary intrusion detection is improved.

The invention discloses a rapid infrared small target detection method and system based on local gray feature discrimination. The method comprises the following steps: receiving infrared image data; gray scale division is carried out on the original gray scale space of the infrared image, and a gray scale histogram is constructed; performing window sliding on the gray histogram to obtain a gray histogram statistical result of a sliding window; obtaining a Gaussian filtering result of the sliding window, completing a target enhancement operation, and obtaining a Gaussian core enhancement value; on the basis of a gray level histogram statistical result, carrying out real-time discrimination on a central pixel of the sliding window, obtaining a target prior response value, and obtaining a target prior mask position map; performing multi-scale contrast calculation on the infrared image to obtain local contrast response values of all potential target points; and taking the pixel position corresponding to the maximum value of the local contrast response value as a target point coordinate, and outputting a detection result. According to the method, the target area can be accurately positioned, and efficient and real-time detection requirements are met.

The invention relates to the field of in-vitro diagnosis, and discloses a peripheral red blood cell image segmentation method based on self-adaption and edge enhancement, and the method comprises the steps: obtaining a to-be-segmented red blood cell image; performing contour enhancement on the to-be-segmented red blood cell image to obtain a target red blood cell image; calculating a gray level histogram of the target red blood cell image, and extracting an envelope line in the gray level histogram; determining two main wave crests on the envelope line and a wave trough between the two main wave crests through a maximum value filtering method; performing threshold segmentation on the target red blood cell image by taking the pixel value corresponding to the trough as a threshold to obtain a binary image; carrying out segmentation point detection and matching on the binary image, and connecting the matched segmentation points to segment independent red blood cells; and performing ellipse fitting on the independent red blood cells to optimize the contour of the red blood cells. According to the method disclosed by the invention, the problem of effectively segmenting the red blood cells is solved, so that the adhered red blood cells and low-pigment red blood cells are effectively segmented out.

The invention relates to the technical field of image processing, in particular to a background grayscale detection method, a storage medium, computer equipment and a program product. According to the background gray level detection method, Gaussian fitting of different Gaussian peak numbers is carried out on a to-be-fitted interval through combination of gray level histogram analysis and Gaussian fitting, one Gaussian curve is selected as a final Gaussian curve, and the minimum mean value of the final Gaussian curve is selected as a central value of a background gray level range. According to the method, the change of gray level distribution in a complex image can be well coped with, the accuracy of the detected background gray level range is ensured, the background gray level range is prevented from deviating towards a high gray level area and interfering other bright parts in the image, automatic detection of the background gray level can be achieved, manual intervention is not needed, and the detection efficiency is improved.

The invention provides a soft tissue adaptive enhancement method and enhancement device for a skull side position film, and the method comprises the steps: calculating a binarization threshold value of a skull side position image according to a gray level histogram and a cumulative distribution function, and generating a binary image of the skull side position film according to the binarization threshold value; extracting a mouth and nose soft tissue region from the binary image; carrying out local self-adaptive dynamic enhancement on the extracted mouth and nose soft tissue area; and carrying out weighted fusion on the enhanced mouth and nose soft tissue area image and the original image of the skull side position film to form a soft tissue enhanced image of the skull side position film. According to the method, the binarization threshold value of the skull side bitmap is calculated by fully utilizing the rule of the brightness range of the mouth-nose soft tissue area, the mouth-nose soft tissue area image can be accurately positioned, and through the local self-adaptive dynamic enhancement and image fusion technology, on the premise that gray level distribution of other anatomical structures is not damaged, the image fusion precision is greatly improved. The visual effect of the mouth and nose soft tissue area image is obviously improved, and more reliable image support is provided for clinical diagnosis.

The invention relates to the technical field of image data processing, in particular to a visual sense-based battery case stamping surface quality detection method, which comprises the following steps of: extracting a collected surface image of a battery case by utilizing semantic segmentation to obtain a battery case area, and equally dividing the battery case area into a plurality of image blocks; for any image block, constructing a gray level histogram of each pixel point; determining the gray level special degree of each pixel point; determining the gray abnormal degree of the target pixel point; obtaining a weighted fuzzy membership degree used for optimizing the fuzzy entropy algorithm; and obtaining a fuzzy entropy value used for determining the burr defect degree of each image block so as to realize quality detection of the stamping surface of the battery shell. According to the method, the gray level of each image block of the battery shell area is analyzed to obtain the gray level special degree and the gray level abnormal degree, so that the fuzzy membership degree between the pixel points is optimized, and the problem that tiny burrs and printing parameter areas are easily confused by a traditional fuzzy entropy algorithm is effectively solved.

The invention relates to the technical field of power transmission line detection, in particular to a nondestructive detection method and system for a power transmission line. The unmanned aerial vehicle is used for putting the airborne nondestructive testing robot to the power transmission line through the special mounting device, and the unmanned aerial vehicle is adaptive to various line structures and the number of sub-wires. And the ground remote workstation drives the robot to move through 5G communication and carries out X-ray detection to obtain a high-resolution image. Based on the normalized matching degree of the X-ray image of the to-be-detected crimping fitting and the corresponding structural feature template, determining an adaptive denoising threshold value of wavelet transform, denoising the X-ray image, and segmenting the X-ray image into a steel core area and an aluminum pipe area through gray histogram analysis; and carrying out local adaptive enhancement by adopting a multi-scale Retinex algorithm, and obtaining a nondestructive testing result of the power transmission line through a multi-task defect identification network fused by attitude correction and a geometric attention mechanism. The nondestructive testing precision of the power transmission line is effectively improved.

The invention discloses an image background region recognition method and device, a storage medium and a product, and relates to the technical field of image processing, and the method comprises the steps: calculating a gray histogram of a to-be-recognized image, carrying out the peak detection of the gray histogram, and obtaining candidate peaks; performing validity judgment on each candidate peak to obtain an effective peak; and determining whether the to-be-identified image has a background area according to the number of the effective peaks and a background determination threshold. Through calculation of a gray level histogram and peak detection, peak value structure characteristics of image gray level distribution can be accurately obtained, a basis is provided for background existence judgment, effectiveness judgment is performed on candidate peaks, false peaks caused by noise interference can be eliminated, a peak value analysis process is more stable, and through comparison of the number of effective peaks and a background determination threshold value, the peak value analysis accuracy is improved. The existence state of the background area in the image is recognized, whether the image contains the background area or not is accurately judged under the scene that the imaging condition changes and the background existence state is uncertain, and the accuracy and robustness of background segmentation are improved.

According to the FPGA-based low-light-level image real-time enhancement method provided by the invention, by utilizing the characteristics of low power consumption, high parallel computing degree and flexible programming of the FPGA, real-time enhancement is carried out on an image acquired by an image sensor in a low-light-level environment, and night target information which cannot be distinguished by human eyes is changed into a visual image. The method comprises the steps of Gaussian filtering, Gamma transformation, image blocking, gray level histogram cutting and reconstruction, histogram equalization and bilinear interpolation on an FPGA (Field Programmable Gate Array). According to the method, the processing speed is higher, the root-mean-square error of the enhanced image is obviously reduced, the peak signal-to-noise ratio is obviously improved, and the structural similarity is doubled.

The invention relates to the technical field of intelligent navigation, and particularly discloses an AR-SLAM intelligent navigation system and method for an underground station building, and the method comprises the following steps: S1, obtaining all underground garages in an urban area, selecting shooting points, and obtaining a two-dimensional image sequence; establishing a virtual three-dimensional model according to the two-dimensional image sequence, and marking the position of the station building to obtain an offline three-dimensional model; s2, converting the shot image into a gray level image, obtaining a gray level histogram, and obtaining an illumination intensity value of the shot image according to the gray level histogram; s3, acquiring a current image in front of the current user, setting standard illumination intensity, and adjusting the illumination intensity values of the current image and the shot image to the standard illumination intensity; comparing to determine a current shooting point where the user is located; and S4, obtaining the position of a target station building, obtaining an optimal path in the offline three-dimensional model according to the current shooting point and the position of the target station building, and prompting the optimal path to a user.

The invention relates to the technical field of image compression, in particular to a grayscale image efficient compression method based on image processing. Dividing image sub-blocks according to the size characteristics of the grayscale image; obtaining a single-peak characteristic image sub-block and a multi-peak characteristic image sub-block according to the pixel point quantity characteristic of the gray level in the gray level histogram of the image sub-block; obtaining different peak domains according to the number characteristics of pixel points of different gray levels of the multi-peak characteristic image sub-blocks; according to the peak spacing feature and the distribution feature between the peak domains, obtaining an overlapping feature value; obtaining a peak value difference degree according to the height difference characteristic between the peak domains; and mapping gray levels of pixel points in the multi-peak feature image sub-blocks according to the overlapped feature values and the peak value difference degree, and re-judging kurtosis attributes of the multi-peak feature image sub-blocks. According to the method, the multi-peak feature image sub-blocks and the single-peak feature image sub-blocks are compressed in different compression modes, so that the compression effect is improved while the image quality is ensured.

The present application belongs to the technical field of image processing, and particularly relates to a radiator weld defect recognition method and system, which comprises the following steps: collecting a radiator image, segmenting a weld area and obtaining a center axis by clustering, position analysis and center axis extraction; in the weld area: calculating the structure deviation and texture difference of each pixel point; determining the defect suspicion degree of each pixel point by fusing the two indexes, counting the gray level histogram of the weld area, including the frequency of each gray level; calculating the defect suspicion degree of each gray level, weighting the frequency of each gray level by using the defect suspicion degree of each gray level; performing histogram equalization based on the weighted frequency, and positioning the defect area by threshold segmentation. The present application changes the image enhancement process from global optimization to focusing on the suspected defect area, thereby improving the recognition accuracy of low-contrast weld defects.

The invention discloses a dynamic target image intelligent analysis system and method based on deep learning, and relates to the technical field of image processing, a visual sensor and an infrared sensor are used for collecting RGB images and infrared images of bird flock respectively, the weights of the RGB images and the infrared images are set respectively, the RGB images and the infrared images are fused, and the dynamic target image intelligent analysis method based on deep learning is obtained. A bionic image is obtained; generating gray level histograms of the target area and the background area for each image block, and calculating contrast entropy of the two gray level histograms; calculating a correction parameter of each frame of bionic image by using the mean value and the standard deviation; performing correction mapping on each frame of bionic image by using the correction parameters, constructing a bionic vision filtering model, training the bionic vision filtering model through deep learning, setting a smoothing function, performing smoothing processing on each frame of filtered bionic image, and splicing all the frames of filtered bionic images to obtain a final bionic image.

The invention discloses a Gaussian point cloud training system and method based on background separation, and relates to the technical field of Gaussian point clouds, and the method comprises the steps: obtaining a historical construction record, extracting a corresponding feature point, obtaining the scene change degree of an image frame set based on the coordinate position of the feature point in a space coordinate system, and obtaining the scene change degree of the image frame set; obtaining the image change degree of the image frame set based on the gray level histogram of the image in the image frame set, and drawing a scatter diagram; according to the method, a current target video is obtained, a target frame number interval is obtained according to the target video and the scatter diagram, key frames and non-key frames are extracted and processed to obtain a rendered static background, and then fusion optimization of a foreground and a background of a scene in the target video is achieved. According to the method, the key frame and the non-key frame are analyzed, so that the problem of efficient training of a dynamic foreground and static background mixed scene is effectively solved, repeated calculation is avoided, the calculation complexity is reduced, and the efficiency is remarkably improved.

The present application relates to the technical field of flaw testing, in particular to a tension clamp forging flaw testing system and method based on optical means, which comprises a vibration detection unit, a fuzzy mapping unit and a segmentation testing unit. The conversion mapping module of the present application performs gray level histogram statistics on the de-fuzzed tension clamp image, calculates the cumulative distribution value of the gray level, linearly maps the cumulative distribution value of the gray level to the target gray level range, generates a new gray level after rounding, realizes image equalization through pixel gray level replacement, enhances the image through image equalization, makes the detailed features on the surface of the tension clamp more obvious in the equalized image, reduces the covering of the detailed features on the surface of the tension clamp by the rough background in the de-fuzzed tension clamp image, increases the difference between the gray level of the detailed features on the surface of the tension clamp and the surrounding rough background area, and uses the gray level to distinguish the detailed features and the rough background to increase the accuracy of the flaw testing.

The invention belongs to the technical field of image processing, and particularly relates to a radiator weld defect identification method and system, and the method comprises the steps: collecting a radiator image, segmenting a weld region through clustering, position analysis and central axis extraction, and obtaining a central axis; in the welding seam area, the structural deviation degree and the texture difference of each pixel point are calculated; the defect suspected degree of each pixel point is determined by fusing the two indexes, and a gray level histogram of a welding seam area is counted, including the frequency of each gray level; calculating the defect suspected degree of each gray level, and weighting the frequency of each gray level by using the defect suspected degree of each gray level; histogram equalization is carried out based on the weighted frequency, and a defect area is positioned through threshold segmentation. According to the method, the image enhancement process is converted from global optimization to focusing on the suspected defect area, and the recognition accuracy of the low-contrast weld defect is improved.

The invention discloses an intelligent image defect repairing method based on deep learning, and relates to the technical field of image processing, and the method comprises the steps: converting an RGB image into an image of a logarithmic space; de-noising processing is carried out on the image, and smoothing processing is carried out on the image by using a filter; simulating a human eye vision system, and separating reflection components in the original image to obtain a reflection component image; carrying out linear normalization processing and nonlinear mapping on the reflection component image; performing color recovery, edge sharpening and contrast improvement processing on the reconstructed image; carrying out remapping on the gray level histogram of the image; and carrying out effect evaluation on the optimized image, adjusting algorithm parameters according to an evaluation result, improving image quality, and obtaining a target image. Through an image enhancement algorithm, a logarithmic space conversion technology, a denoising and smoothing processing technology and a histogram equalization technology, an original RGB image is optimized and repaired, so that a user obtains better visual experience.

The invention relates to the technical field of machine vision detection, and discloses a programmed automatic exposure method and system for machine vision long-distance displacement monitoring, and the method comprises the steps: obtaining a template picture corresponding to each target, and calculating a dark space ratio; the method comprises the following steps: capturing a camera video stream, performing standardized scaling on an image region of interest, and generating a gray histogram through transverse splicing or weighted accumulation; analyzing the grey level histogram, determining a dark area end point and a dark and dark area peak value based on a dark area proportion, calculating a boundary score and a double-peak difference value score, and weighting to obtain a final score; and determining an optimal exposure value and an optimal gain value step by step to form a program type automatic exposure parameter. The system corresponds to the method. By adopting the method and the device, the dynamic automatic optimization of the exposure parameters is realized, the stable imaging quality of the target is ensured, and the precision and the reliability of machine vision long-distance displacement monitoring are effectively improved.

The application discloses an image enhancement method and device, relates to the technical field of image enhancement processing, and is used for improving the effect of image enhancement. The method comprises the following steps: acquiring a gray level histogram of a luminance component of a current image frame; grouping the gray level histogram according to a human eye luminance sensitivity curve to obtain at least two histogram groups; the at least two histogram groups comprise a first histogram group and a second histogram group, the maximum gray level value corresponding to the first histogram group is smaller than the minimum gray level value corresponding to the second histogram group; determining a target stretching relationship according to stretching steps of the at least two histogram groups and the gray levels corresponding to the at least two histogram groups; the target stretching relationship is used for representing the relationship between the gray level corresponding to the current image frame and the gray level corresponding to a target image frame; the target image frame is an image frame obtained after the current image frame is enhanced; and the target image frame is obtained based on the target stretching relationship.

The application relates to the technical field of image processing, in particular to a fresh flower image segmentation method based on a finite element method, which comprises the following steps: collecting a fresh flower image by customizing an optical scheme; pre-processing the collected fresh flower image; performing image segmentation on the pre-processed image by the finite element method; realizing feature extraction on the segmented image by using a traditional gray level histogram method; and performing classification by using an SVM algorithm. The application better depicts the gray level change of a point neighborhood, so that the boundary information can be better segmented, and the accuracy of segmenting a certain region into regions containing completely different property features in the fresh flower sorting process is improved.

The embodiment of the application discloses a kind of magnetic resonance image gray histogram matching method and device, first, establish multiple image data sets, respectively according to the image of all samples in each image data set, the average gray histogram of corresponding image data set is generated.For each image data set, Gaussian process regression model between imaging parameter combination and average gray histogram is established.Then, the image to be matched and the physical parameter data and imaging parameter combination data of the image to be matched are obtained, the Gaussian process regression model corresponding to the image to be matched is determined according to the physical parameter data of the image to be matched.The imaging parameter combination data of the image to be matched is substituted into corresponding Gaussian process regression model, and the approximate gray histogram of the image to be matched is generated.Finally, a preset template histogram is obtained, and the gray level mapping function of the image to be matched is established using the approximate gray histogram and the template histogram.The gray level mapping function is used to correct the gray level of the image to be matched.

The application provides a license plate intelligent recognition method and system based on image enhancement, and belongs to the technical field of image processing. The method performs Fourier transform on license plate image data to determine the spectral distribution characteristics; then, a gray level histogram is constructed based on the characteristics, and when the gray level variance exceeds the preset variance threshold, it is determined that the license plate is affected by light changes and the dynamic characteristic index of the gray level distribution is determined; then, the index is analyzed using a convolutional neural network to generate modulation parameters of the contrast enhancement factor and the sharpening intensity; when the deviation between the spectral peak position and the preset peak value exceeds the preset deviation threshold, it is determined that the license plate is affected by motion blur, and an enhanced spectral representation is constructed; finally, the enhanced license plate image data is generated through inverse Fourier transform. The quality of the license plate image in a complex environment is significantly improved, the characters are clear and identifiable, and the license plate recognition accuracy and system robustness are effectively improved.

The invention relates to an underwater image intelligent real-time enhancement method and apparatus, and an electronic device. The method comprises the steps of obtaining a to-be-processed underwater image; rapid degradation analysis is carried out on the underwater image, a low-dimension statistical feature set is extracted, and the low-dimension statistical feature set is composed of a whole-frame gray level histogram mean value, a local contrast space mean value and a red-blue channel gray level difference value; comparing the low-dimension statistical feature set with a preset threshold set, and determining a degraded scene category of the underwater image; according to the degradation scene category, internal processing parameters of a lightweight Transform enhancement engine are dynamically modulated, and the internal processing parameters comprise feature fusion weights in a multi-scale feature fusion decoder; and performing enhancement processing on the underwater image by using the enhancement engine after parameter modulation, and outputting an enhanced image. According to the method, real-time processing of high-resolution images can be realized under the constraint of finite computing power, and various underwater image degradation scenes including composite degradation can be effectively identified and adaptively processed.