55 results about "Image averaging" patented technology

The invention discloses an improved YOLOv11-based high-reflection and high-transmittance material surface defect detection method, which comprises the following steps: S1, acquiring a mobile phone screen surface defect image, preprocessing and labeling, and generating a training data set containing three defects of scratches, edge breakage and cracks; s2, constructing a defect detection model; s3, averagely dividing an input image into four independent detection regions, adopting a bounding box overlapping area weight distribution strategy for cross-region defects, and performing partition counting on the image by using a defect detection model; s4, constructing a surface defect detection network meeting the real-time requirement of the industrial production line, and outputting a detection result and a processing frame rate; s5, pixel-level mask segmentation is carried out on the scratch defect, and contour coordinates and geometric features are extracted; and S6, fitting the actual damage size of the scratch defect on the surface of the high-reflection and high-transmittance material according to the segmentation result. According to the method, the detection speed is high under the condition that the accuracy is ensured, and a reliable solution is provided for surface flaw detection of the high-reflection and high-transmittance material.

The invention discloses an SLAM front-end optimization method based on brightness grading and gradient constraint, and relates to the technical field of computer vision and vision SLAM. In order to solve the defects that adaptive grading enhancement and parameterization control based on illumination types are lacked in the prior art, and feature point quality, spatial distribution uniformity and front-end real-time performance are difficult to consider in a high-frequency input scene, the invention provides a comprehensive brightness grading and feature optimization scheme. The method comprises the following steps: firstly, calculating the average brightness of an input image, dividing the image into a dark light type, a normal type and an overexposure type, and respectively adopting gamma correction, contrast limited adaptive histogram equalization and inversion enhancement strategies for different types to realize illumination adaptive enhancement; then screening high-quality feature points with significant local curvature changes; and updating the detection area. According to the method, the feature stability, the matching precision and the real-time performance of the SLAM front end in a complex indoor environment are remarkably improved, and the method is suitable for a self-localization and mapping system.

The invention discloses a running water image inspection method, device and equipment and a medium. The method comprises the following steps: acquiring a to-be-processed pipeline image and inspection reference data corresponding to the to-be-processed pipeline image; averagely segmenting the to-be-processed pipeline image into N sub-images, and inputting the N sub-images into the image feature extraction model to obtain a feature map of the to-be-processed pipeline image; inputting the test reference data into the data feature extraction model to obtain data features of the test reference data; inputting the feature map and the data features into a pipeline image classification model to obtain a category label of the to-be-processed pipeline image; the classification label is real or forged; and according to the category label, determining an inspection result of the to-be-processed pipeline image. According to the embodiment of the invention, the method can automatically, comprehensively and accurately inspect the pipeline image based on the feature map of the pipeline image and the data features of the inspection reference data which can be used for auxiliary inspection, and determines whether the pipeline image is a real pipeline image or a forged pipeline image.

A surface defect detection method for optically detecting a surface defect in a strip-shaped body includes an image acquisition step of detecting reflected light from the strip-shaped body obtained by illuminating a surface of the strip-shaped body and imaging while relatively scanning the surface of the strip-shaped body to acquire a plurality of images including the surface of the strip-shaped body, an average image calculation step of calculating an average image of the acquired images, an image correction step of performing shading correction on each acquired image using the average image to obtain corrected images, and a defect detection step of detecting a surface defect in the strip-shaped body based on the corrected images. The average image calculation step includes recognizing an inspection target region in which the strip-shaped body is located in each image and contributing to the average image only for pixels in the inspection target region.

A processing method of a display screen brightness compensation includes: obtaining a sample data set of the display screen, a real environment temperature and a real image average grayscale value corresponding to the real environment temperature, the sample data set including an environment temperature sample set, a image average grayscale sample set corresponding to the environment temperature sample set and a simulated temperature sample set; performing a linear regression processing on the sample data set to generate a regression coefficient set; processing the real environment temperature and the real image average grayscale value according to the regression coefficient set to generate a temperature to be compensated of the display screen; and determining a brightness compensation parameter according to the temperature to be compensated, and performing a brightness compensation processing on the display screen according to the brightness compensation parameter.

This invention belongs to the field of image processing technology and discloses a method, system, and terminal for continuous image target detection considering relative target motion. It utilizes an image target detector to detect targets in an image at time k, obtaining the corresponding detection result Y. k ; Regarding the detector's detection result Y k Y k Temporal matching is performed with the detection results obtained by the algorithm at the previous time step; based on the average brightness index g of the image at time k. k Estimate the reliability parameter p of the test results k Calculate the noise matrix R(k) and the predicted bounding box size b″ of the detection results at time k, respectively. k The Kalman filter is applied to the position of the object at time k, and the bounding box size of the object at time k is estimated by weighted fusion to form the result of image detection at time k. This invention can estimate the motion parameters and size variation patterns of targets in images based on target detection results in continuous images, providing effective data support for subsequent target localization and actual motion state estimation.

This method for acquiring and processing images of a scene (4) provided by a detection system (2) comprising a sensor (6) comprises the modification of an adjustment parameter of the sensor (6), the modification of a reference offset table of the sensor (6) applied to the images and the compensation of a variation in the average intensity of the images generated by the modification of the adjustment parameter of the sensor (6) and of the reference offset table of the sensor (6) applied to the images. Figure for the abstract: Fig 2

This invention provides a scene-adaptive fast automatic exposure adjustment method, comprising: S1. Acquiring a frame image and the exposure parameters of the previous frame, denoted as E_pre; S2. Calculating the average brightness Y_cur, the proportion of overexposed pixels P_sat, and the proportion of underexposed pixels P_dark of the current frame image using the current frame image; S3. Calculating the slope between the target brightness and the average brightness of the current frame, denoted as k1: k1=Y_target / Y_cur; S4. Determining whether the current brightness meets the expectation. If k1_th_dark≤k1≤k1_th_sat, where k1_th_dark is the lower limit of the k1 tolerance threshold and k1_th_sat is the upper limit of the k1 tolerance threshold, the exposure parameters remain unchanged, and the next loop is performed, returning to step S1; otherwise, if the expectation is not met, the exposure parameters are updated and the process proceeds to step S5; S5. Selecting the corresponding convergence parameters; S6. Calculating and acquiring the exposure parameters, and updating the exposure parameters. This method can quickly achieve automatic exposure control in bright, dark, or scenes with drastic brightness changes, allowing the image to converge to a reasonable brightness quickly.

The application discloses a method for judging and eliminating ground light in high-resolution images, and belongs to the field of space visible light detection, and the steps are as follows: firstly, the input visible light image is evenly divided into four areas according to rows, and the brightness sum of each area is calculated; secondly, a ground light judging factor is calculated, and if the ground light judging factor is greater than 0, it is indicated that the ground light exists; and then, the brightness of the image with the ground light is balanced according to rows, so that the purpose of eliminating the ground light is achieved.

The invention relates to the technical field of optical system design, in particular to a design method of a multi-scale convolutional network and single-plane diffraction element optical number cooperation system, and aims to solve the problems that an existing diffraction calculation imaging technology is poor in restoration real-time performance and low in restoration precision, the method comprises the following steps: step 1, designing a single-plane diffraction element; 2, constructing a multi-scale convolutional network image restoration module; and step 3, constructing an image restoration loss function. According to the invention, a high-quality color image is obtained while a unit piece system is extremely simplified. An MSE loss function considering full-view-field wavefront aberration and diffraction efficiency weight of RGB three channels is constructed, and the restoration effect of the design method is improved by 80% compared with the average PSNR of an image before restoration; compared with a traditional PSF deconvolution image restoration algorithm, the average PSNR is improved by 30%, and a new thought is provided for miniaturization development of an optical system in the future.

The application discloses a method for solving color deviation of a three-color Mini LED backlight liquid crystal display device and a liquid crystal display device, and relates to the technical field of liquid crystal display, and comprises the following steps: adjusting RGB backlight currents of a white field, recording a set of RGB backlight current ratios when color coordinates of the white color reach design target values; adjusting RGB backlight currents of red, green and blue single-color fields respectively, recording a second set of RGB backlight current ratios, so that the RGB synthesis obtains the design target values, and obtaining coordinate points W, R, G and B of the white color, the red color, the green color and the blue color on a CIE chromaticity diagram respectively; performing segmentation processing on a to-be-displayed image, counting average image color coordinates of each block as to-be-displayed colors, obtaining a coordinate point C of the to-be-displayed colors on the CIE chromaticity diagram, and calculating RGB current values of backlight of the block according to relative positions of C, R, G, B and W and the two sets of RGB backlight current ratios. The application can guarantee a color gamut and solve the color deviation phenomenon.

The invention provides an image enhancement method, device and equipment, which can enhance effective information in an image. The image enhancement method can be applied to the image enhancement device. The image enhancement device can obtain a to-be-processed image, then a maximum pooling feature image is obtained according to the to-be-processed image, a first area in the maximum pooling feature image corresponds to a second area in the to-be-processed image, and the value of pixels in the first area is the maximum value in the values of multiple pixels in the second area. The image enhancement device can also obtain an average pooling feature image according to the maximum pooling feature image, a third region in the average pooling feature image corresponds to a fourth region in the maximum pooling feature image, and the value of a pixel in the third region is an average value of values of a plurality of pixels in the fourth region. In addition, the image enhancement device can also obtain a target image according to the to-be-processed image and the average pooling image, and the target image comprises less noise information compared with the to-be-processed image.

The invention provides an image brightness adjusting method and device, electronic equipment and a storage medium, and relates to the technical field of image processing, and the method comprises the steps: obtaining an input image, and determining the image average brightness of the input image; when it is detected that the image average brightness is smaller than a brightness lower limit value, the external light intensity of the input image is determined; and performing gamma correction on the image brightness of the input image under the condition that the external light intensity is smaller than an external light intensity threshold value. According to the invention, the brightness characteristic of the image and the external light environment factor are effectively combined, so that the targeted brightness adjustment is realized, and the picture distortion caused by invalid operation and improper adjustment in a scene without adjustment is avoided.

The application provides a two-dimensional grating preparation method and a diffractive grating waveguide device, and the two-dimensional grating preparation method comprises the following steps: determining a setting scheme of two grating vectors of a two-dimensional grating based on two known first basic grating vectors and second basic grating vectors of the two-dimensional grating; determining grating design optimization parameters of the two-dimensional grating, including a grating period and a grating line direction, according to the setting scheme of the two grating vectors of the two-dimensional grating; and preparing the two-dimensional grating according to the grating design optimization parameters of the two-dimensional grating. The two-dimensional grating preparation method provided by the application is helpful to realize the balance between the coupling-out image energy efficiency and the average brightness uniformity of the coupling-out image of the two-dimensional grating, and improve the degree of freedom of the two-dimensional grating preparation process.

The present invention discloses an adaptive repeated scanning method and apparatus. The method includes: scanning a target area corresponding to a target feature at a current scanning position to obtain a first image set of the target feature at the current scanning position; performing a multi-image averaging operation on all the scanned images included in the first image set to obtain an initial image; determining whether there are abnormal features in the initial image; when it is determined that there are abnormal features in the initial image, repeatedly scanning the target area according to a determined additional scanning number to obtain a second image set; performing a multi-image averaging operation on all the scanned images included in the first image set and all the scanned images included in the second image set to obtain a target image. It can be seen that implementing the present invention can obtain more scanned images by increasing the scanning number to obtain a target image with higher imaging quality when and only when there are abnormal features in the initial image.

The invention provides an exposure parameter adjusting method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a license plate image from a received original image, and dividing the license plate image into a plurality of image blocks in response to the situation that the environment brightness information of the original image is smaller than a preset brightness threshold value, the plurality of image blocks comprise a target image block and a non-target image block, the first weight of the target image block is greater than the second weight of the non-target image block, determining the target image block corresponding to the license plate area from the plurality of image blocks, and determining the average brightness of the license plate image based on the plurality of image blocks, the first weight and the second weight, and in response to the condition that the average brightness is greater than the overexposure threshold, smoothly adjusting the image exposure parameter based on the exposure adjustment increment, thereby realizing division of the target image block and the non-target image block from the license plate image, reducing the interference of the non-license plate image area on the exposure brightness adjustment of the license plate image area, and improving the image brightness adjustment accuracy. Therefore, effective exposure brightness adjustment of the license plate image is completed.

The invention relates to an automatic exposure method for internal imaging of an aluminum casting hole in an industrial quality inspection scene in the technical field of industrial imaging. The automatic exposure method comprises the following steps: determining the maximum exposure time corresponding to an image average gray value meeting a detection requirement; determining an exposure time sampling interval; and performing interpolation polynomial fitting according to the maximum exposure time and the exposure time sampling interval. Compared with a conventional mode of determining exposure time in industrial AOI quality inspection, the method can save a lot of manpower work; the detection step length and the exposure time are decoupled, so that the method is more flexible; the method can adapt to the influence of the inner wall roughness and the bottom roughness caused by the abrasion of a machining tool on imaging to a certain extent.

The invention relates to a nondestructive testing technology of electrical equipment, and discloses an infrared detection device for cracks of an electrical tube, which comprises a support plate, a closed detection camera obscura, a rotary moving mechanism, an infrared imaging unit and an active infrared lighting unit, and is also provided with an infrared light source intensity real-time adjusting system, the system integrates an environment state evaluation module, a system state evaluation module, an adaptation evaluation module and a motion state evaluation module, and can sense multi-dimensional parameters such as environment light intensity, air suspended matter concentration, light source working temperature, optical window transmissivity, production line operation speed and pipeline rotation angular velocity in real time; and the infrared illumination intensity is dynamically adjusted by integrating the average gray scale of the image and the target feature contrast ratio. According to the invention, intelligent self-adaptive control of the infrared light source in a complex and changeable production environment is realized, the image quality stability and the accuracy and reliability of crack detection are effectively improved, and the method is suitable for full-automatic online nondestructive detection in large-scale production of power tubes.

The present application relates to the technical field of optical coherence tomography, and discloses a retinal OCT full-automatic imaging and image enhancement method, which comprises the following steps: automatically positioning the three-dimensional space position of the pupil through a binocular vision system and aligning an imaging light beam; cooperatively determining an optimal refractive compensation value based on the joint evaluation of the interference signal intensity and the axial registration stability of a short sequence image; collecting multiple B-scan images, performing motion correction, removing false frames and image averaging to generate a final image; performing quality evaluation on the final image, and adaptively retriggering the acquisition process or re-focusing according to the evaluation result. The present application realizes the automation of the whole imaging process by introducing cooperative refractive compensation and closed-loop feedback control, enhances the anti-interference ability to the motion of the measured person, reduces the dependence on the experience of the operator, and can stably obtain high signal-to-noise ratio retinal tomographic images.

The design method of the optical-digital collaborative system of a multi-scale convolutional network and a single-plane diffraction element relates to the field of optical system design technology. In order to solve the problems of poor real-time restoration and low restoration accuracy of existing diffraction computational imaging technology, the method includes the following steps: Step 1, designing a single-plane diffraction element; Step 2, constructing a multi-scale convolutional network image restoration module; Step 3, constructing an image restoration loss function. The present invention achieves high-quality color images while achieving extreme simplification of the single-element system. An MSE loss function of the RGB three-channel considering the full-field wavefront aberration and diffraction efficiency weights is constructed. The restoration effect of this design method is improved by 80% compared with the average PSNR of the image before restoration; compared with the traditional PSF deconvolution image restoration algorithm, the average PSNR is improved by 30%, which also provides a new idea for the future miniaturization of optical systems.

This invention discloses a two-photon microscopy imaging method and apparatus for simultaneous multi-region aberration correction, belonging to the field of microscopic imaging. The apparatus includes an aberration control module, a multifocal generation module, a scanning imaging module, an imaging objective, a fluorescence collection module, and an aberration correction module. The aberration correction module tests multiple aberration modes and fits the amplitude of the selected aberration mode based on the functional relationship between the average image intensity and the amplitude of the aberration mode. Finally, a corresponding compensation phase is applied to the phase modulator to achieve simultaneous multi-region aberration correction. This apparatus, through multifocal scanning and multifocal aberration control, achieves simultaneous aberration correction in multiple imaging regions, improving the speed of aberration correction and realizing aberration correction for large field-of-view dynamic signals.

The invention discloses an anti-counterfeiting code image scatter reflection detection method based on top-hat operation. The method comprises the following steps: 1) obtaining an original anti-counterfeiting code image sample through image acquisition equipment; 2) positioning and extracting a target image anti-counterfeiting code area in the sample, and carrying out image preprocessing; 3) binarizing the target image and performing top-hat operation, and extracting astigmatism points in the image; 4) calculating the image average gradient of the image, and judging whether the image is scatter reflection; and calculating the average gradient of the scatter result image of the top-hat operation, comparing the average gradient with a set gradient threshold value, and judging that the corresponding scatter point is scatter reflection if the average gradient exceeds the threshold value. According to the method disclosed by the invention, when scattered light reflection occurs in the image, the scattered light reflection can be effectively and accurately detected, the replacement of a shooting anti-counterfeiting code can be prompted, the false alarm phenomenon of anti-counterfeiting detection caused by blurring is reduced, the stability of an anti-counterfeiting detection algorithm is improved, and the robustness of identification is greatly improved.

The invention discloses a multi-terminal collaborative road illumination comprehensive evaluation method and system. The method comprises the steps of constructing a street lamp geographic coordinate database; acquiring street lamp illuminance values, image gray features and timestamps collected by detection equipment on different vehicles; identifying and rejecting the interference spectrum to obtain detection data of multiple vehicles without interference; on the basis of a correlation coefficient between the street lamp illuminance value and the image average gray value and a spectrum relative coefficient corresponding to the street lamp illuminance value, verifying the detection data of the multiple vehicles subjected to interference removal so as to determine effective detection data of the multiple vehicles; determining the weight of the effective detection data of the multiple vehicles based on the street lamp geographic coordinate database; combining the effective detection data of the multiple vehicles with the corresponding weights to determine a comprehensive illumination value; and determining the state of the street lamp according to comparative analysis of the comprehensive illumination value and historical data so as to update a street lamp state file. By implementing the method provided by the invention, accurate monitoring and early warning of performance degradation of the street lamp can be realized.

The present invention discloses a grating imaging processing method, device, system, and medium. The method includes: when the moiré fringes on the detector are uniformly distributed, obtaining a projection image of at least one angle of a sample placed in the grating imaging system, as well as a background image when no sample is placed; combining the projection image with the background image to separate absorption information, refraction information, and scattering information, and simultaneously obtaining an absorption information image corresponding to bright fringes in the moiré fringes, an absorption information image corresponding to dark fringes, an average absorption information image, a refraction information image, and a scattering information image. Thus, the grating imaging system can obtain multimodal imaging, and the sample structure can be analyzed based on the images of each modality, thereby obtaining more comprehensive information about the sample's interior.

The invention relates to a control method for prolonging the service life of a liquid crystal screen of a projector, which comprises the following steps of: adjusting the brightness of a blue light source and a response curve of the transmittance of a blue light liquid crystal screen corresponding to the blue light source according to the image average brightness of a blue digital image in input digital images, and the transmittance value of the response curve of the transmittance of the blue light liquid crystal screen is increased along with the reduction of the brightness of the blue light source. The method for controlling the long service life of the liquid crystal screen of the projector has the advantages that the service life of the liquid crystal screen is longer, and color cast of a displayed picture is avoided.

The invention relates to the technical field of image color enhancement, in particular to a color correction method for an ultra-high-definition camera, and the method comprises the steps: obtaining each frame of image in video data; averagely dividing each frame of image into a plurality of areas; obtaining a target existence coefficient of each region in each frame of image according to the color difference between different color blocks in each region, the proportion of edge pixel points and the dispersion degree of the average gray value of all edges, and obtaining a target existence coefficient of each region in each frame of image by combining the total number of edges of each region in an adjacent region in each frame of image and the color difference and texture difference in all adjacent frames; the method comprises the following steps: acquiring a correction coefficient of each region in each frame of image, further judging whether each region of each frame of image needs to be segmented into small regions, and when segmentation is needed, acquiring side lengths of the small regions after segmentation of each region in each frame of image, and further acquiring a video subjected to color correction. According to the invention, the region where the target object is located in each frame of image is divided into smaller regions, so that the color correction effect of the video is improved.

This invention provides a method and system for evaluating the smoke extraction effect of a range hood, relating to the technical field of equipment detection. The method includes: after the range hood starts operating, acquiring a target image set, wherein the target image set includes instantaneous images captured by a camera within a preset time period and under conditions free from stray light interference at a preset frequency; the instantaneous images are images captured when a parallel light source of preset thickness and wavelength illuminates the oil fume thermal plume generated by the smoke-generating device; converting the instantaneous images into grayscale images; processing the grayscale images using an image averaging algorithm and an image standard deviation algorithm to determine the probability density map of the oil fume concentration distribution and the oil fume boundary distribution map of the oil fume thermal plume; and determining the smoke extraction effect of the range hood based on the probability density map of the oil fume concentration distribution and the oil fume boundary distribution map of the oil fume thermal plume, thus solving the technical problem of poor accuracy in existing range hood smoke extraction effect evaluation results.

The invention relates to the technical field of ferrous metallurgy, and discloses a vision and pressure fused intelligent detection method for the depth of an RH dip pipe inserted into molten steel. The method comprises the following steps: when an RH vacuum tank is vacuumized, starting a built-in camera to collect and preprocess a tank bottom image signal to obtain an RH tank bottom preprocessed image; extracting an average pixel gray value and a local maximum brightness gradient of each frame of image as key features; constructing a two-dimensional feature image based on the key features, constructing a model by using an artificial intelligence algorithm, and determining the moment when the molten steel overlays the bottom of the RH vacuum tank; and reading the vacuum pressure at the moment, and calculating the depth of the dip pipe inserted into the molten steel by combining the preset length of the dip pipe, the molten steel density, the slag density and the slag thickness. Through visual feature extraction and vacuum pressure data fusion, the depth of the dip pipe inserted into the molten steel in the RH production process is accurately calculated in cooperation with the neural network model, and the problem that depth detection is inaccurate due to the fact that traditional detection is prone to being disturbed by the molten steel and slag is solved.

A method for automatically selecting time points for multi-target ISAR imaging includes the following steps: S1: performing pulse compression on a broadband echo sequence for multi-target imaging; S2: performing range image averaging and calculating the waveform entropy of the average range image; S3: taking the time corresponding to the range image with the maximum waveform entropy as the time reference and performing a Keystone transform; S4: performing azimuth compression on the range image sequence to obtain a multi-target range-Doppler two-dimensional image; S5: after target separation on the range-Doppler plane, performing motion compensation and R-D imaging processing on each target to obtain an ISAR image of each target. By calculating the waveform entropy of the average range image, the present invention can automatically select the time point at which the multi-target overlap is minimized, thereby improving computational efficiency. The automatically selected time point exhibits the best separability of target echoes, thereby improving imaging efficiency.

The invention provides a highlight target identification method based on a dung beetle optimization algorithm and an adaptive threshold segmentation algorithm, and the method comprises the steps: carrying out the Gaussian filtering of continuous frames of active and passive images to remove a part of noise, and carrying out the morphological opening operation to reduce the noise and maintain the feature size of a target; then registering the images, eliminating the image average gray difference of the active image and the passive image caused by the influence of active light illumination by using image difference operation, and preliminarily extracting a possible target area; then optimizing the maximum between-cluster variance by using a dung beetle optimization algorithm to determine a difference image segmentation threshold value, binarizing the difference image, and further extracting a target candidate region; for the candidate target area, the highlight target is identified by adopting comprehensive shape measurement of circularity and eccentricity, and the method has higher detection accuracy; the efficient search mechanism of the dung beetle optimization algorithm meets the real-time requirement, and the processing speed is higher; the method is suitable for actual scenes with uneven illumination and complex noise, and has higher robustness.