256 results about "Image histogram" patented technology

An adaptive image tone mapping curve based on perceptual preference guidelines is generated as a sigmoidal function, in which the sigmoidal function parameters (slope and shift) are determined by original image statistics. Tone curves generated for different images each have a smooth sigmoidal shape, so that the tone mapping process does not change the image histogram shape drastically. The sigmoidal function has the form: <math-cwu id="MATH-US-00001"> <number>1< / number> t it ( x ) = 100 1 + exp it ( - alpha it ( x / 100 - beta ) ) , <mathematica-file id="MATHEMATICA-00001" file="US20020171852A1-20021121-M00001.NB" / > <image id="EMI-M00001" wi="216.027" he="18.96615" file="US20020171852A1-20021121-M00001.TIF" imf="TIFF" ti="MF" / > < / math-cwu> where alpha is the slope parameter and beta is the shift parameter. The input value x in the sigmoidal function varies in the range [0, 100], because the tone curve is generated on an L* scale, which has values from 0 to 100. The sigmoidal tone curve calculation can be implemented efficiently using simple arithmetic operations by pre-calculating and storing various factors used in the calculation of alpha and beta and by pre-generating a pair of fixed tone curves with two extreme slopes and interpolating between the curves.

Adaptive vision-based vehicle detection methods, taking into account the lighting context of the images are disclosed. The methods categorize the scenes according to their lighting conditions and switch between specialized classifiers for different scene contexts. Four categories of lighting conditions have been identified using a clustering algorithm in the space of image histograms: Daylight, Low Light, Night, and Saturation. Trained classifiers are used for both Daylight and Low Light categories, and a tail-light detector is used for the Night category. Improved detection performance by using the provided context-adaptive methods is demonstrated. A night time detector is also disclosed.

Adaptive vision-based vehicle detection methods, taking into account the lighting context of the images are disclosed. The methods categorize the scenes according to their lighting conditions and switch between specialized classifiers for different scene contexts. Four categories of lighting conditions have been identified using a clustering algorithm in the space of image histograms: Daylight, Low Light, Night, and Saturation. Trained classifiers are used for both Daylight and Low Light categories, and a tail-light detector is used for the Night category. Improved detection performance by using the provided context-adaptive methods is demonstrated. A night time detector is also disclosed.

An image desired to be reproduced is scanned to determine its video pixel grey values. A histogram generator generates a histogram distribution representing a frequency of the grey values. The histogram distribution is analyzed to determine a minimum and maximum input grey values which define input boundaries. A segment point is computed between the input boundaries based on the histogram data. The segment point defines a plurality of input segments between the input boundaries. A dynamic output range is selected. Each input segment is mapped to an output segment based on a linear transformation for the corresponding segment. In this manner, a tone reproduction curve map having a piecewise linear transformation is automatically generated from the image histogram data.

An image retrieval system and method that combines histogram-based features with Wavelet Frame decomposition features, as well as two-pass progressive retrieval process. The proposed invention is robust against illumination changes as well as geometric distortions. During the first round of retrieval, moment features of image histograms in the Karhunen-Loeve color space are derived and used to filter out most of the dissimilar images. During the second round of retrieval, multi-resolution WF decomposition is recursively applied to the remaining images. A set of coefficients of low-pass filtered subimages at the coarsest level, after being mean-subtracted and normalized, are utilized as features containing spatial-color information. Modulus and direction coefficients are calculated from the high-pass filtered X-Y directional subimages at each level, and central moments are derived from the direction histogram of the most significant direction coefficients to obtain TRSI direction / edge / shape features. Since the proposed invention is fast and robustness against illumination and geometric distortions, the invention is quite appealing for real-time image / video database indexing and retrieval applications.

An apparatus, method, computer useable medium, and processor programmed to automatically generate tone mapping curves in a digital camera based on image metadata are described. Rather than having a static tone mapping curve for all images, the curve can be varied automatically based on, e.g., the brightness histogram of the image. In one embodiment, a certain percentage of the least bright pixels and a certain percentage of the brightest pixels can be disregarded, while the remaining pixels can be linearly stretched to encompass the original range of brightness values. Based on the distribution of the resultant stretched brightness histogram, slopes for the low end (S0) and high end (S1) of the tone mapping curve can be independently determined, and the tone mapping curve can be automatically generated. The improved, automatically generated tone mapping curve may be able to lift shadows more aggressively and / or increase the dynamic range of the image.

An adaptive image tone mapping curve based on perceptual preference guidelines is generated as a sigmoidal function, in which the sigmoidal function parameters (slope and shift) are determined by original image statistics. Tone curves generated for different images each have a smooth sigmoidal shape, so that the tone mapping process does not change the image histogram shape drastically. The sigmoidal function has the form: t⁢(x)=1001+exp⁢(-α⁢(x / 100-β)),where α is the slope parameter and β is the shift parameter. The input value x in the sigmoidal function varies in the range [0, 100], because the tone curve is generated on an L* scale, which has values from 0 to 100. The sigmoidal tone curve calculation can be implemented efficiently using simple arithmetic operations by pre-calculating and storing various factors used in the calculation of α and β and by pre-generating a pair of fixed tone curves with two extreme slopes and interpolating between the curves.

The invention is directed to methods of detecting and / or measuring a feature in a retinal mage. The feature detected and / or measured may be one or more of the optic disc, optic disc centre, optic disc radius, vessel edge, vessel calibre / width and vessel central reflex. One method for detecting the optic disc includes analyzing an image histogram to determine intensity levels; analyzing the intensity levels to determine a threshold intensity for potential optic disc regions; determining the number of pixels for each potential optic disc region; and calculating the centre of each potential optic disc region from the number of pixels in each potential optic disc region to thereby detect the optic disc.

The invention provides an adaptive extraction method of remote sensing image thematic information based on spectrum matching degree. The method comprises the following steps: firstly, artificially selecting a sample region of a thematic ground object, carrying out PPI calculation in the sample region, selecting pixels with maximum purity as end members of the thematic ground object, extracting spectrums of the end members, and subsequently, utilizing whole matching degree indexes which integrate the similarity of distances and shapes among the spectrums to calculate the matching degree of the spectrums between the pixels and the end members on an image; secondly, adaptively determining the segmentation threshold of a spectrum matching degree image by using an image histogram automatic segmentation method, so as to obtain initial separation of the thematic ground object and the whole background; and next, carrying out statistic segmentation on the gray level mean value and standard deviation of thematic ground object pixels, adaptively determining the threshold of region growth, traversing the thematic ground object pixels and taking the thematic ground object pixels as seed points to carry out region growth; and finally, in a grown thematic ground object local unit, adopting an iteration method to carry out adaptive local classification to continuously approach a real thematic ground object boundary, thereby obtaining a final refined ground object extraction result.

The invention relates to a human face image fusing method based on laplacian-pyramid. On a basis of the laplacian-pyramid decomposition, the invention applies a partial binary mode arithmetic operator into a fusion rule. As for infrared and visible-light human face images in each layer of the laplacian-pyramid, average gradient of the image, standard deviation of an image histogram and Chi square statistic similarity measurement are used for determining the selection of respective weighting coefficients of the infrared and visible-light human face images, and effective fusion is performed by utilizing the determined weighting coefficients so as to obtain a fused image which is rich in detail information, clear and stable and facilitates the observation of human eyes. Compared with visible-light images and near-infrared images, the fused information provided by the invention has greater comentropy but smaller average cross-entropy and mean square root cross-entropy and has more remarkable effect of image fusion.

The invention discloses an improved HEVC sample point self-adaption compensation method. The BO algorithm and the EO algorithm are adopted for performing SAO filtering on images. The sample point self-adaption compensation algorithm is adopted for solving ring effects caused by transform and quantization in the current video coding standard HEVC in the loop filter technology. The current video coding standard HEVC comprises BO and EO algorithms, but sideband distribution conditions are not fully taken into consideration in the BO algorithm which only transmits four continuous sidebands. According to the invention, by use of an image histogram, the sideband distribution conditions can be sufficiently analyzed; intra-frame prediction mode type texture information is extracted in a self-adaption manner; and for different sizes of CTB, four kinds of BO filter modes are added, BO filter pattern discrimination process is improved, and the improved self-adaption compensation method is provided. The experimental results show that compared with the HM10.0, coding performance is improved; bit rate gain of chromaticity can reach -7.53%; and subjective assessment quality of image edge parts is obviously improved.

The invention discloses an image de-spin and stabilization method based on subarea matching and an affine model, which comprises the steps that 1, a reference frame and a current frame are subjected to image histogram equalization processing; 2, four areas distributed crosswise and an image center area are selected as matching areas of motion estimation; 3, an affine transformation model is established, and a current frame motion vector is solved; and 4, a kalman filter model is established, and motion compensation of a current frame displacement vector is achieved. In Step 2, the operand is less; a speed is high; the motion estimation capacity is favorable when translation and rotation motion coexist in an image; and an interframe estimated error is less than one pixel. According to the method, the random motion quantity of a video image can be removed with high precision; scanning motion inherent in a platform or a load is kept; and the smoothness of a dynamic image sequence is improved. The method can effectively improve the video image and inhibit random noise.

A method of segmenting a radiographic image into diagnostically relevant and diagnostically irrelevant regions comprising: acquiring a digital radiographic image including a matrix of rows and columns of pixels; detecting the initial background left point of a histogram of the image; detecting the foreground of the image; regenerating the background of the image by region growing; validating the background of the image; merging the background and foreground regions of the image as diagnostically irrelevant regions; and extracting the anatomy region of the image as the diagnostically relevant region.

The present invention relates to a method for quantization of a color histogram bin value of an image or video, and more particularly, to a method for non-uniform quantization of a color histogram bin value of an image (or video) according to the frequency of color occurrence. The present invention has effects of effectively representing the characteristics of the color histogram of the image better as compared to the conventional art, and improving the performance of the image retrieval when the image (video) retrieval search is conducted.

The invention discloses a lossless data hiding method based on difference image histogram cycle spinning. The lossless data hiding method includes steps of encrypting, secret key transmitting and data recovery. During encrypting, an original image is divided into a plurality of sub image blocks at the transmitting end, cycle spinning operation is performed by aid of a sub block difference histogram to achieve step-by-step embedding of secret data, a secret image C' is formed and transmitted to the receiving end. During secret key transmitting, the size of sub image blocks, the length of embedded data and the embedding level L serve as a secret key to be distributed to the receiving end. During data recovery, the receiving end extracts the secret data step by step from the secret image C' in an inverse method of the embedding process, block difference and original image data are recovered, and secret data transmission and lossless recovery based on the image carrier are achieved. The lossless data hiding method overcomes the defect that the capacity of the original histogram spinning algorithm depends on a single peak point, and increases the embedding capacity through a cycle embedding method. Experimental results prove that the lossless data hiding method can well meet requirements for high capacity, low distortion and high efficiency.

The invention discloses a gamma correction and smoothing filtering based image histogram equalization enhancing method. The method comprises performing gamma correction on original histograms to control the problem of over-high peak values in the original histograms; performing sliding window smoothing filtering on gamma corrected histogram to eliminate mutation in the histograms; applying a traditional histogram enhancing method on the basis of correction of the histograms to obtain target enhanced images. The method has the advantages that the balance is enhanced: balance enhancing can be performed on all portions of the images effectively, and the 'white washing' effect produced due to excessive enhancement can be prevented effectively; image characteristics are maintained effectively: images can be enhanced efficiently, image detail information and average brightness can be maintained, and brightness saturation, brightness great changing and detail loosing are prevented. By the aid of the method, the images can be enhanced in a high quality mode.

A method for histogram calculation using a graphics processing unit (GPU), comprises storing image data in a two-dimensional (2D) texture domain; subdividing the domain into independent regions or tiles; calculating in parallel, in a GPU, a plurality of tile histograms, one for each tile; and summing up in parallel, in the GPU, the tile histograms so as to derive a final image histogram.

The invention discloses a self-adaptive image contrast enhancement method based on histograms, which is used to judge whether image gray scales are concentrated based on the total number of pixels corresponding to several continuous gray scales on contraction histograms and traverse histograms. For images with concentrated gray scales, namely the images with the total number more than a threshold, contrast enhancement operations cannot be performed so as to avoid image quality transformation after enhancement. Then the minimum key gray scale, a mid-value key gray scale and a maximum key gray scale are obtained through transformation based on a minimum gray scale, a maximum gray scale and a gray scale average value obtained by the contracted histograms. Finally, space mapping relationships are established based on four spaces divided by the minimum key gray scale, the mid-value key gray scale and the maximum key gray scale for the minimum gray scale, the gray scale average value and the maximum gray scale, a lookup table for image contrast enhancement is obtained, and image contrast is enhanced for input images based on the lookup table. Therefore, excessive enhancement of image contrast is avoided based on distribution conditions of image histograms and self-adaptive regulation mapping relationships.

The invention discloses a digital video image enhancement achieving system and method based on an FPGA. The digital video image enhancement achieving method based on the FPGA mainly comprises the steps of dehazing processing on a current frame image and anti-reflection processing on the current frame image, wherein the step of dehazing processing on the video image is achieved through a dark channel prior dehazing algorithm, and a good dehazing effect can be achieved; the step of anti-reflection processing on the current frame image is achieved through an image histogram equalization algorithm, and the effect of converting a low-illumination-level gray image into a clear image can be achieved. By the adoption of the image dehazing process and the image anti-reflection process, the brightness, contrast ratio and color saturation of the image can be improved, and the image can be clearer, transparent and saturated in color.

The invention aims to provide a Tetrolet transform-based multichannel satellite cloud picture fusing method, which comprises the following steps of respectively performing image histogram equalization processing on multichannel cloud pictures to be fused, and respectively performing Tetrolet transform to obtain a low frequency coefficient, a high frequency coefficient and mosaic coverage values corresponding to the low frequency coefficient and the high frequency coefficient; in a low frequency part of a Tetrolet domain, decomposing again by using a Laplacian pyramid, taking a mean value for a top layer, and reconstructing after taking large parts of grey absolute values of other layers; in a high frequency part of the Tetrolet domain, taking a larger part of high frequency coefficient standard deviation in each image mosaic, and taking corresponding values as the mosaic coverage values; obtaining a final fused image through Tetrolet inverse transformation. Experimental results show that the method can be well used for fusing multichannel cloud pictures, the fused image is good in visual effect and can clearly retain typhoon eye and cloud system detail information, a typhoon center is positioned at high accuracy by using a fused result, and the method is suitable for typhoons with eyes and non-eye typhoons.

The present invention provides an image adaptive enhancement method based on histogram compactness transformation. The method comprises a step 1 of gathering the gray scale frequency histograms of the input images, and calculating the characteristic values of the image histograms; a step 2 of calculating a search starting point and the left and right reference frequencies; a step 3 of carrying out the histogram compactness transformation; a step 4 of establishing a gray scale mapping relation; a step 5 of carrying out the input image gray scale reconstruction. The method of the present invention considers the own characteristics of the input images, so that the insufficiencies of the enhancement methods, such as a classic linear stretching transformation method, a histogram truncation stretching method, etc., are overcome effectively. Experiments show that the method is obvious in enhancement effect, has good adaptability and practicality, can be widely used for the enhancement of various low contrast images and the high bit wide quantization image-to-low bit wide quantization image conversion processing, and is a rapid and effective image automatic enhancement method.

The invention discloses an H.264 code rate control method based on frame complexity, belonging to the technical field of video encoding. The method comprises the following steps of: A. determining the target bit number of a current encoding frame according to the frame complexity; and B. calculating quantitative parameters and carrying out rate-distortion optimization. The H.264 code rate control method is characterized in that the frame complexity is judged through a complexity coefficient, wherein the complexity coefficient is defined as the sum of an average difference value of an image histogram and a weighted average absolute error ratio of an image brightness component, and the traditional target bit formula is amended through an empirical formula on the basis of the complexity coefficient. The invention can not only control the code rate more accurately, but also improve the average peak signal-to-noise ratio of a video output sequence, thereby obtaining the better image quality. The invention simultaneously reduces the quality fluctuation among video image frames and ensures that the video image quality is smoother.

The invention provides a front face feature-based vehicle type recognition method, which comprises the following parts: S01, executing an image histogram information-based road surface vehicle automatic extraction method: analyzing road surface images sent back by a traffic checkpoint on a road, and extracting possible vehicle areas in the road surface images by adopting a monocular image analysis method; S02, executing a color and gradient information-fused vehicle front face interception method: analyzing the color and the gradient information of a target in vehicle area images obtained in the step S01 to complete the interception of a vehicle front face; S03, performing heterogeneous sample analysis-based vehicle type online training, and establishing vehicle templates of various vehicle types; S04, executing a vehicle front face feature subspace-based vehicle type judging method: matching the vehicle front face intercepted in the step S02 and the vehicle templates obtained in the step S03 to obtain the judging decision of the vehicle types. According to the front face feature-based vehicle type recognition method disclosed by the invention, the automatic recognition of the vehicle types can be accurately performed, and the daily work of relevant departments requiring vehicle type information is greatly facilitated.

An image processor computes a raw histogram for an unprocessed raw digital image. The image processor positions a clip point at an intensity level within the raw image histogram. The pixels having an intensity level less than the intensity level of the clip point are blacked-out (reduced in intensity). The remaining illuminated pixels are displayed.

A method of segmenting regions of an image wherein a number of partitions are determined based on a range of an image histogram in a logarithmic luminance domain. Regions are defined by the partitions. A mean value of each region is calculated by K-means clustering wherein the clustering is initialized, data is assigned and centroids are updated. Anchor points are determined based on the centroids and a weight of each pixel is computed based on the anchor points.

A system and method are disclosed for detecting and labeling places recognized in a video stream using change-points detection. The system includes a segmentation module and a label learning module. The segmentation module is configured to receive a video stream comprising multiple digital representations of images. The video stream is represented by a measurement stream comprising one or more image histograms of the video stream. The segmentation module segments the measurement stream into multiple segments corresponding to place recognized in the videos stream. The segmentation module detects change-points of the measurement stream and computes probability distributions of the segments over multiple pre-learned place models. The label generation module is configured to generate place labels for the places recognized by the place models.

The three-sided rolling fingerprint acquiring and splicing process includes acquiring background image, acquiring serial fingerprint, eliminating background from the fingerprint image, mapping the image histogram to form histogram with homogeneously distributed grey scale, Gaussian smoothing treatment, calculating threshold and binarizing the image with the threshold, judging whether to have fingerprint in the image and acquiring the next image if no or progressing the next step if yes, calculating the center and four boundaries of the fingerprint, acquiring the next sequential image and repeating the forgoing steps, obtaining the maximum correlating block between the two sequential fingerprint image and its center line as the optimal splicing line, eliminating dislocation in the splicing line to splice two fingerprint images, repeating the foregoing steps until no fingerprint input.

The invention belongs to the technical field of image processing and in particular relates to a self-adaptive image segmentation method based on a fuzzy threshold value. The self-adaptive image segmentation method comprises the following steps: step 1, pre-processing a histogram to acquire an image histogram with double-peak properties; step 2, carrying out gradient detection on the pre-processed image histogram and determining the position of a wave trough; step 3, determining the position of a wave peak according to the position of the wave trough; step 4, determining the distance between two adjacent wave peaks according to peak values of the wave peaks; calculating window width sizes of membership functions of different images according to the distances between the different wave peaks; and step 5, determining a segmentation threshold value. The self-adaptive selection of the window width is realized, and the disadvantage that the segmentation of images on the histogram with inconspicuous double peaks cannot be easily realized by the fuzzy threshold value is effectively improved, an applicable range of the fuzzy threshold value image segmentation method is expanded, and the segmentation effect of the fuzzy threshold value segmentation method is improved.

An apparatus and method of determining a convergence angle of a stereo camera. The method includes setting interest regions by first and second cameras, respectively, the interest regions having a same size and being symmetric to each other; photographing images by the first and second cameras while varying a convergence angle of the stereo camera by a predetermined degree for each photographed image of the first and second cameras, respectively; analyzing image histograms of the interest regions for each of the first and second cameras; and setting an optimum convergence angle.

The invention discloses an eyeground image blood vessel segmentation method based on self-adaption difference of Gaussians. The segmentation method comprises steps of 1) extracting colorful eyeground image green channels, and performing pretreatment of self-adaption histogram equalization and anisotropy coupling diffusion with limited contrast ratio; 2) constructing Gaussian scale space; 3) subtracting adjacent two layers in the Gaussians scale space so as to get difference of Gaussian images; 4) averaging difference of Gaussian image weighing so as to get blood vessel increasing images; 5) performing binaryzation for the blood vessel increasing images; 6) rotating Gaussian kernel in 12 directions, wherein 15 DEG is regarded as step size in 0-180 DEG, repeating steps of 2-5, and overlapping results in the 12 directions; 7) selecting 20% of the grey value of the second peak as a light area of a threshold value extracting images according to bimodality of a pretreatment image histogram; and 8) reducing light areas from the blood vessel binary images so as to reduce effects of the light areas on segmentation of blood vessels. The method is widely applicable for blood vessel segmenetaion of all kinds of colorful eyeground images.