2377 results about "Gradation" patented technology

A system and process for broadcast and communication with bi-level or sketch video at extremely low bandwidths is described. Essentially, bi-level and sketch video presents the outlines of the objects in a scene being depicted. Bi-level and sketch video provides a clearer shape, smoother motion, shorter initial latency and cheaper computational cost than do conventional DCT-based video compression methods. This is accomplished by converting each color or gray-scale image frame to bi-level or sketch image frame using adaptive thresholding method, compressing bi-level or sketch image frames into bi-level or sketch video using adaptive context-based arithmetic coding method. Bi-level or sketch video is particularly suitable to such small devices as Pocket PCs and mobile phones that possess small display screen, low bandwidth connection, and light computational power.

The present invention relates to devices and methods for the measurement and / or for the specification of the perceptual intensity of a visual image, or the perceptual distance between a pair of images. Grayscale test and reference images are processed to produce test and reference luminance images. A luminance filter function is convolved with the reference luminance image to produce a local mean luminance reference image. Test and reference contrast images are produced from the local mean luminance reference image and the test and reference luminance images respectively, followed by application of a contrast sensitivity filter. The resulting images are combined according to mathematical prescriptions to produce a Just Noticeable Difference, JND value, indicative of a Spatial Standard Observer, SSO. Some embodiments include masking functions, window functions, special treatment for images lying on or near borders and pre-processing of test images.

The invention discloses a workpiece apparent defect detection method based on machine vision. Firstly, a visual system is used for guiding a robot, the pose of a target workpiece is precisely positioned according to a template matching algorithm based on a gray value, and then workpiece apparent defect detection is carried out and comprises the steps that firstly, a workpiece image is acquired and pretreated through median filtering; secondly, the target workpiece is divided through a global threshold value, and workpiece pose correction is carried out; thirdly, burr interference of the edge of the workpiece is removed through mathematical morphology open operation; fourthly, notches, material sticking, cracking, indentation, needle eyes, scratches and foaming apparent defects are detected. The method solves the problems that the artificial detection speed is slow, the efficiency is low and precision is poor; the problems that according to current vision detection, defect types are singular, the imaging quality is poor, and the false drop rate is high are solved, and the automation degree of precise workpiece production and product quality are improved.

The invention relates to the technical field of digital image processing and pattern recognition, discloses a fuzzy clustering steel plate surface defect detection method based on pre classification and aims to overcome defects of judgment missing and mistaken judgment by the existing steel plate surface detection method and improve the accuracy of steel plate surface defect online real-time detection effectively during steel plate surface defect detection. The method includes the steps of 1, acquiring steel plate surface defect images; 2 performing pre classification on the images acquired through step 1, and determining the threshold intervals of image classification; 3, classifying images of the threshold intervals of the step 2, and generating white highlighted defect targets; 4, extracting geometry, gray level, projection and texture characteristics of defect images, determining input vectors supporting a vector machine classifier through characteristic dimensionality reduction, calculating the clustering centers of various samples by the fuzzy clustering algorithm, and adopting the distances of two cluster centers as scales supporting the vector machine classifier to classify; 5, determining classification, and acquiring the defect detection results.

The invention discloses a fast monocular vision odometer navigation and positioning method fusing a feature point method and a direct method, which comprises the following steps: S1, starting the vision odometer and obtaining a first frame image I1, converting the image I1 into a gray scale image, extracting ORB feature points, and constructing an initialization key frame; 2, judging whether thatinitialization has been carry out; If it has been initialized, it goes to step S6, otherwise, it goes to step S3; 3, defining a reference frame and a current frame, extracting ORB feature and matchingfeatures; 4, simultaneously calculating a homography matrix H and a base matrix F by a parallel thread, calculating a judgment model score RH, if RH is great than a threshold value, selecting a homography matrix H, otherwise selecting a base matrix F, and estimating a camera motion according to that selected model; 5, obtaining that pose of the camera and the initial 3D point; 6, judging whetherthat feature point have been extracted, if the feature points have not been extracted, the direct method is used for tracking, otherwise, the feature point method is used for tracking; S7, completingthe initial camera pose estimation. The invention can more precisely carry out navigation and positioning.

The invention relates to a tunnel lining crack width measuring method, which comprises the follow steps of: (1) picking up crack images through a digital camera and measuring pickup distance through a laser distance measuring instrument; (2) selecting a region required to be measured on the picked images and converting selected region images into gray level images; (3) determining the threshold value of the gray level images selected in the second step, cutting binary image cutting, and obtaining a target region; (4) extracting a white region only comprising crack in the target region; (5) extracting the edge of the crack by a sub pixel method, and obtaining sub pixel level edge images; (6) circulating the pixel width of the crack by a minimum distance method; (7) circulating the actual width of the obtained crack according to the calibration proportion of the predetermined actual pixel size and the image pickup distance; and (8) storing the obtained actual width data into a crack data base. Compared with the prior art, the tunnel lining crack width measuring method has the advantages of low cost, high speed, convenience, high precision and the like.

The invention relates to a process for the registration of two different images of the same object consisting of breaking down each of the images into space components representing the distribution of the grey levels of the image, applying a phase registration method to the components to bring about a correspondence between the components of one image with those of the other image, summating all the results of the bringing into correspondence and detecting, in the image resulting from said sum, the maximum grey level defining the transformation between the two initial images.

The invention provides an eye precision orientation technology in the field of human face identifying. It is characterized in that it provides an eye precision orientation method on complex background image. It adopts a micro structure character with high efficiency and high redundant to express eye mode local and area grey distributing character and uses AdaBoost computing method to choose the most differential micro structure character to form the strong classer.

The invention discloses a real time panorama video splicing method based on ORB characteristics. The real time panorama video splicing method based on the ORB characteristics comprises the following steps: acquisition of multipath synchronized video data is started; pretreatment is carried out on images in various paths at a same moment, and color images are changed into gray scale images of 256 levels, and a de-noising processing is carried out on the images through employing a Gaussian filter; the ORB feature extraction algorithm is employed to carry out feature point extraction on the images in the various paths at the same moment, and ORB characteristic vectors of the feature points are calculated; through the adoption of the nearest neighborhood matching method and the RANSAC (random sample consensus) matching method to determine a homography matrix array between corresponding frames of the synchronized videos; frame scene splicing is carried out according to the homography matrix array; and finally spliced videos are output. The real time panorama video splicing method based on ORB characteristics and the apparatus are advantageous in that: the feature extraction speed and the coupling effect are improved in the image splicing process.

The present invention relates to a method of processing a video image sequence in a liquid crystal display panel and to a device for implementing this method. According to one embodiment of the invention, at least one motion-compensated image is generated for each group of m consecutive images of the sequence, m being greater or equal to 2, in order to obtain a group of n consecutive images, with n>m. Said group of n consecutive images replaces the group of m consecutive images into the sequence. Next, for each pixel having in a current image of the new sequence a current grey level and in the next image a different target grey level, an intercalary grey level is calculated which is higher or lower than said target grey level depending on whether said target grey level is respectively higher or lower than the current grey level of the pixel. Next, in the current image, the current grey level of the pixels is replaced with said calculated intercalary level. This method makes it possible to correct the blurring effects due to the mode of display and to the high response time of the LCD panel.

This invention provides an image processing device in which a pixel satisfying a certain unwanted gradation condition with respect to all the plurality of color components is judged as a unwanted pixel derived from unnecessary image data, and each of gradation data of the plurality of color components composing the unwanted pixel is converted into a certain background gradation which is predetermined with respect to each one of the color components. The image processing device eliminates a possibility that a pixel satisfying the unwanted gradation condition with respect to part of the color components is judged as the unwanted pixel. This arrangement enables to securely extract and remove the unwanted pixel attributed to the unnecessary image data such as undertone and backside image data and prevents degraded image reproduction due to color change resulting from erroneous removal of part of the color components of the necessary image data.

The invention discloses an image enhancement and partition method, which comprises the following steps of I, image enhancement: a processor and an image enhancement method based on fuzzy logic are adopted for performing enhancement processing on an image needing to be processed, and the process is as follows: i, converting from an image domain to a fuzzy domain: mapping the grey value of each pixel point of the image needing to be processed to a fuzzy membership degree of a fuzzy set according to a membership function (shown in the specifications); ii, performing fuzzy enhancement processing by utilizing a fuzzy enhancement operator in the fuzzy domain; iii, converting from the fuzzy domain to the image domain; II, image partition: partitioning a digital image subjected to enhancement processing, i.e. a to-be-partition image, according to an image partition method based on a three-dimensional fuzzy division maximum entropy. The method is simple in steps, reasonable in design, convenient to realize, good in processing effect and high in practical value, and an image enhancement and partition process can be simply, conveniently and quickly completed with high quality.

The present invention performs gradation conversion linearly on an image signal when the image signal is darker than a preset value. Consequently, the dark region of the image is not distorted nonlinearly, and the gradations of shadows and hair can be reproduced in full richness. On the other hand, gradation conversion is performed nonlinearly on image signals if the image signal is brighter than a preset value. The nonlinear conversion in this case: (1) has a slope that effectively equalizes an average noise amplitude of the image signal without being based on the output gradation value, and (2) is offset so as to be continuous with the gradation conversion characteristic of the dark area gradation conversion unit. Through condition (1), the level-dependent noise in the bright region is equalized. Furthermore, through condition (2), gradation differences and gradation reversals do not occur at the boundary between the dark region and the bright region.

A dynamical image of is an array of black-and-white images, or frames, of arbitrary dimension. Dynamical images are constructed from gray scale and color images, video sequences etc. A method of topological analysis and decomposition of dynamical images through computation of homology groups of the frames is provided. Each frame is partitioned into a collection of components, which, in turn, have tunnels, voids, and other higher dimensional cycles. The cycles in each frame are linked to the cycles in each adjacent frame to record how they merge and split. Further, the dynamical image is simplified by removing from frames all cycles that are small in terms of length, area, volume, etc, or lifespan. Applications of the method lie in image enhancement and restoration, motion tracking, computer vision, surface and curve reconstruction, scientific image analysis, image recognition and matching.

An image processing system includes a local area selecting section configured to select a local area as at least one correction target from an input image, the local area including at least one pixel; a correction amount calculating section configured to calculate a gradation correction amount for the local area as the correction target; and a local gradation correcting section configured to perform a gradation correction to correct a gradation value of the local area based on the gradation correction amount to generate an image with a local gradation correction. Also, the correction amount calculating section includes a peripheral area gradation value calculating section configured to calculate a gradation value of a peripheral area of the local area; and a local area correction amount converting section configured to calculate the gradation correction amount by using the calculated gradation value of the peripheral area.

The invention discloses a method for extracting a text in a complex background image. The method comprises the steps of firstly, using the susan operator to detect and identify angular points in a source image, after removing isolated angular points, conducting integral projection transformation to cut out suspected text regions, and screening and removing non-text regions according to priori knowledge; then, judging the background complexity of a text region by the utilization of gray level jump information, when a background is judged to be complex, conducting color clustering on the text region by the utilization of the kmeans clustering algorithm, and determining the type which a text belongs to and extracting the text according to the color information at the position where the angular points are the densest; when the background is judged to be simple, conducting binaryzation on the image by the utilization of a largest-between-class variance method; finally, realizing accurate extraction of the text region. The text extraction method can locate the text region in the complex background image, and finally extracts characters after removing the background.

The invention puts forward an image registration method based on an inertial sensor and a camera. The camera and the inertial sensor are synchronously controlled to operate to simultaneously acquire the color image and depth image of the same scene and the three-axis motion angular velocity and three-axis motion acceleration of the camera; Kalman filtering is utilized to carry out filtering on the three-axis motion angular velocity and the three-axis motion acceleration, the Eulerian dynamics equation is used for converting the three-axis acceleration into a Eulerian angle, the uniformly accelerated motion formula is used for converting the three-axis angular velocity into three-axis displacement, and thereby an external reference matrix of the camera is obtained; by utilizing the affine transformation relation between the color image and the depth image and the proportional relation between depth image grey value and pixel depth information, the depth information of angular points in the depth image is solved; an OFCM (Orthogonal Frequency-division Multiplexing) model is used for calculation to obtain image background optical flow; the RANSAC (Random Sample Consensus) algorithm is used for completing image registration. The calculation load of the method is low, both the real-timeness and the precision are high, and the method is applicable to hand-held devices and mobile devices.

The invention discloses an obstacle avoidance robot based on visual feature binding and a reinforcement learning theory. Visual feature binding and the reinforcement learning theory are utilized, and depth image information is integrated to learn priori knowledge of complex environment on the basis of a convolutional network and image matching technology; on the basis of a feature binding mechanism, a result is fed back to obstacle avoidance of the obstacle avoidance robot to make a decision intelligently, quickly, efficiently and accurately; an MNIST handwriting dataset is adopted, different control signal meanings are given to numbers 0-9 respectively, a Kinect sensor is utilized to collect RGB images with moving distance containing handwriting pictures in the complex environment, and an image matrix is created and output through gray-scale processing; image matching is performed, pictures after being matched are subjected to binarization processing and substituted into a well-trained image recognition model based on a convolutional neutral network, numbers in front of the sensor are judged finally, and corresponding control signals are transmitted to a robot control module through a host to realize intelligent obstacle avoidance.

The invention brings forward a rapid and automatic mosaic technology of aerial video in a search and tracking system. The technology provided by the invention comprises the following steps: two continuous frames of color images in an aerial video sequence are converted to grayscale images; three-valued processing is carried out on the grayscale images by an image three-valued method, edge feature information images are extracted by an edge detection sobel operator and feature points are extracted by an Harris algorithm; feature point matching is completed by an optical flow pyramid algorithm; optimal homography matrix of the color images is calculated by combining a DLT algorithm and a RANSAC algorithm; panoramic image mosaics is completed after image correction of the color images by correction geometric transition matrix; and dynamic regulation of image three-valued upper and lower thresholds and a threshold of the edge detection sobel operator is carried out during the process of panoramic image mosaics. According to the invention, image mosaic accuracy and robustness are raised.

The invention provides a pretreatment method of iris images based on gray information. A rough circle center (xo, yo) of a pupil is located through binary, mathematical morphology, gray projection and other operations, the point, which has a gray value larger than T and the shortest distance to the (xo, yo), is searched at a plurality of lines close to the (xo, yo) from the center to the two sides as a pupil boundary point; the accurate circle center and the radius of the pupil can be located through curve fitting; the gray level first difference of the lines where the searched pixel points lie is calculated and the point with the largest value of the sum of the level first difference is searched in the range of places where iris external boundary point possibly appears and taken as the iris external boundary point; the accurate circle center and radius of the iris external boundary is obtained after curve fitting. The quality of iris image is judged through calculating point sharpness and number of effective pixel points according to the normalized image. The pretreatment method of iris images of the invention not only can decrease the time needed by repeated iteration during location, but also can judge the quality of iris image accurately and quickly according to the extracted iris image.

A video signal processing apparatus processes an input video signal for displaying an image based on the video signal. Detected is a gradation level of the input video signal. A color saturation level of the input video signal is depressed in a predetermined color saturation level range only when the detected gradation level is located in a predetermined gradation level range. A first input video signal having a first gradation level (or number of bits) is converted into a second video signal having a second gradation level (or number of bits) that is lower (or smaller) than first gradation level (or number of bits), for displaying an image based on the input video signal. Error data is generated in response to a data portion of the first input video signal. The data portion corresponds to a difference between the first and the second gradation levels (or number of bits). The error data is obtained by multiplying the data portion by predetermined error diffusion coefficients according to pixel dots that surround a pixel dot composed of R(red)-, G(green)- and B(blue)-signal components of the first input video signal. At least one of the number of bits of the error diffusion coefficients for one of the R-, G- and B-signal components is different from the other number of bits of the error diffusion coefficients for the other signal components The generated error data is added to the first input video signal to convert the first input video signal into the second video signal. Reverse-gamma correction may be applied to the first input video signal before conversion.

The invention relates to a pipeline defect identification method based on computer vision and machine learning. The method comprises the following steps: carrying out gray scale stretching on pipelineimages and videos collected by a pipeline robot; performing sliding filtering processing to filter noise interference; extracting the outline of the suspected defect area by adopting a Canny operator; dividing the acquired image into gradients or edge direction histograms of all pixel points in a cell unit, and combining the histograms to extract HOG features of the image; training a VGGNet modelby adopting a transfer learning strategy, and extracting image features by adopting a trained VGGNet feature extractor; fusing the HOG features of the pipeline image and the features extracted by theVGGNet, classifying the features by using an SVM, and identifying the pipeline defects; and counting the pipeline defect identification result and outputting the pipeline defect type. Defect characteristics of a real pipeline environment can be recognized, robustness and recognition accuracy of a classification model are improved, system image processing speed and recognition efficiency are improved, and the method is suitable for the field of urban pipeline and industrial pipeline maintenance.

The invention relates to a measuring method of a shield tunnel lining segment seam open width and an apparatus thereof. The method comprises the following steps: using a computer vision kit to demarcate cameras, collecting segment seam images through the digital cameras and determining a shooting distance through a laser range finder; converting the collected images into grayscale images, selecting an appropriate filtering parameter and using a Gauss filter to smooth the images in a selected region of interest (ROI); determining a maximum threshold and a minimum threshold after the images aresmoothed and carrying out Canny lag threshold segmentation so as to obtain a sub-pixel linear characteristic of the images and a segment seam edge; screening needed sub-pixel segment seam edge imagesfrom the sub-pixel linear characteristic; using a normal vertical minimum distance method to respectively calculate a largest width and a mean width according to the obtained two sub-pixel segment seam edge images; determining a conversion proportion between a pixel dimension and an actual dimension according to a predetermined internal parameter of the cameras and calculating so as to obtain an actual width of the segment seam. The method and the apparatus have advantages of low costs and high precision. The method can be used rapidly and conveniently.

The invention belongs to the technical field of machine vision, and particularly relates to a vision-based method of detecting defects on the end surface of a bearing finished product, aiming at solving the problem of on-line detection of defects such as rust, scratches, pressed marks and bumps which probably exist on the end surface of the bearing finished product. The method comprises the following steps of: acquiring an original acquired image; acquiring a first workpiece image of the end surface of the bearing to be detected; performing edge extraction, obtaining the precise positioning ofthe bearing workpiece by using the fitting algorithm of a circle, and obtaining a second workpiece image; and based on the second workpiece image, expanding an annular workpiece drawing into a rectangular workpiece drawing; dividing the rectangular workpiece drawing into a plurality of end surface region drawings by adopting a unidirectional projection segmentation method; carrying out binarization segmentation through an adaptive gray threshold to obtain a defect candidate region; and as to the defect candidate region, obtaining a detect region by using a connected domain analysis method according to the defect position and shape information. The method realizes the on-line detection of the bearing end surface, improves the detection efficiency and the detection stability.

The invention relates to a Fourier descriptor and gait energy image fusion feature-based gait identification method. The method comprises the steps of performing graying preprocessing on a single frame of image, updating a background in real time by using a Gaussian mixture model, and obtaining a foreground through a background subtraction method; performing binarization and morphological processing on each frame, obtaining a minimum enclosing rectangle of a moving human body, performing normalization to a same height, and obtaining a gait cycle and key 5 frames according to cyclic variation of a height-width ratio of the minimum enclosing rectangle; extracting low-frequency parts of Fourier descriptors of the key 5 frames to serve as features I; centralizing all frames in the cycle to obtain a gait energy image, and performing dimension reduction through principal component analysis to serve as features II; and fusing the features I and II and performing identification by adopting a support vector machine. According to the method, the judgment whether a current human behavior is abnormal or not can be realized; the background is accurately modeled by using the Gaussian mixture model, and relatively good real-time property is achieved; and the used fused feature has strong representability and robustness, so that the abnormal gait identification rate can be effectively increased.

The invention belongs to the digital image processing field and relates to a processing method for retaining a ghosting-free moving object in high-dynamic range image fusion. The objective of the invention is to remove ghosting and retain a high-exposure degree moving object in an image fusion process. The method includes the following steps that: images are converted into gray-scale images, and histogram equalization is carried out on the series of gray-scale images; with a moving object adopted as a singular point, a modeling result of a moving object-removed background image is obtained by utilizing median filtering; the background image is subtracted from the images comprising the moving object, so that images which only comprise the moving object after the background is removed can be obtained; and through morphological operations comprising erosion and dilation, precision processing is performed on the moving object images, large values are assigned for weight portions corresponding to the pixels of the moving object to be retained by using the coordinate information of the pixels of the moving object to be retained, and an obtained result is adopted as a weight map, and then, fusion is performed on the images according to the weight map. The processing method of the invention is mainly applied to digital image processing.

An automated method of improving digital color images at high speed, which supports pipe-lining and has very little memory requirements, and is therefore specially suitable for on-the-fly processing of real time video, as well as for processing of large batches of images without the need of human intervention. The method includes a novel dynamic range adaptation scheme that operates on the norm of the image, which is suitable as is for gray-scale images. For, color images, a simple color reconstruction stage is added that maintains optimal color fidelity.

An image processing apparatus including a specific image detection unit detecting an area including at least a part of a specific image in an input image, a representative color calculation unit calculating a representative color of the specific image on the basis of pixels belonging to the area detected by the specific image detection unit, plural correction curve acquisition units which acquire plural correction curves resulting in gradation correction for every element color on the basis of the gradation value for every element color constituting the representative color after correction, and plural correction units which correct the gradation value for every element color in each pixel constituting the input image based on the obtained plural correction curves.

The invention relates to the technical field of artificial intelligence, in particular to a mechanical part stress corrosion detection method and system based on image processing. The method comprises the following steps: obtaining an initial grayscale image of a mechanical part according to a weighted average method, identifying the defect probability of each pixel point to obtain a first detection result image, extracting an uncertain region in the first detection result image, and adaptively dividing the uncertain region into a plurality of first regions; extracting a neighborhood feature value of the pixel points of the first region, and obtaining a second gray level image according to a weighted average method; obtaining a second gradient according to the neighborhood feature value, obtaining an optimal statistical feature having the smallest difference with the standard statistical feature according to the second gradient, and further obtaining an optimal coefficient; and obtaining an optimal grayscale image by adopting the optimal coefficient, and obtaining a final detection result image according to the optimal grayscale image. Through the feature enhancement of the image, the detection accuracy of the stress corrosion area is improved, and the detection efficiency is also improved while the detection accuracy is improved.

The invention relates to a camera deformation calibration method and quick calibration method based on predicted template. Based on the unchanged linear perspective projection to calculate the ideal image coordinate to the analog linear distance and the minimum target, predicting the lens deformation coefficient through nonlinear optimization function. Based on camera deformation coefficient, it predict the deformation coordinate of the ideal image coordinate and stores it into the template matrix, directly searching the template reading ideal image corresponding deforming coordinate, completing the quick calibration of deformed image through linear insert value to get the gray value of the ideal image pixel. The put forward deformation coefficient calibration method is low in cost, simple in algorithm, and easy realization.