113 results about "Color quantization" patented technology

A method, system, and computer-readable storage medium for applying Gaussian Mixture Models (GMMs) to local image patches using an adaptive color lookup table. Per-channel color quantization may be performed to find representative colors for a local image patch. Each combination of the representative values corresponds to a representative color. The probabilities of the representative colors may be computed using a local GMM color model and stored to corresponding entries in an adaptive color lookup table. For every pixel in an image patch, the closest representative color may be found, and the corresponding probability may be retrieved from the lookup table and used for the pixel. The method may, for example, be applied to each local window in a method for automatically determining segmentation in a digital video image sequence to calculate the foreground probabilities for the pixels in a propagated classifier via a GMM.

A method for detecting an object in a digital image including the steps of performing color quantization on a model image including the object and on a search image that potentially includes the object, generating a plurality of search windows, computing spatial-color joint probability functions of each model and search image, where the color co-occurrence edge histogram is chosen to be the spatial-color joint probability function, assessing the similarity of each search image to the model, and designating search windows as containing the target object.

An image processing system includes an image encoder system and a image decoder system that are coupled together. The image encoder system includes a block decomposer and a block encoder that are coupled together. The block encoder includes a color quantizer and a bitmap construction module. The block decomposer breaks an original image into blocks. Each block is then processed by the block encoder. Specifically, the color quantizer selects some number of base points, or codewords, that serve as reference pixel values, such as colors, from which quantized pixel values are derived. The bitmap construction module then maps each pixel colors to one of the derived quantized colors. The codewords and bitmap are output as encoded image blocks. The decoder system includes a block decoder. The block decoder includes a block type detector, one or more decoder units, and an output selector. Using the codewords of the encoded data blocks, the comparator and the decoder units determine the quantized colors for the encoded image block and map each pixel to one of the quantized colors. The output selector outputs the appropriate color, which is ordered in an image composer with the other decoded blocks to output an image representative of the original image. A method for encoding an original image and for decoding the encoded image to generate a representation of the original image is also disclosed.

The invention discloses a self-adaptation camouflage design method based on a background. According to the method, on the basis of a texture synthesis algorithm based on pixels, a whole camouflage image is synthesized pixel by pixel; in the synthesis process, a self coding algorithm is adopted to extract a texture primitive containing background sample complete texture features, and a neighborhood of the pixels to be synthesized is determined; by means of the neighboring correlation principle and a helix supplement search mechanism, a neighborhood matched with a target pixel neighborhood is searched for in a background sample; similarity matching is carried out through a pixel and texture similarity matching rule, pixels in the neighborhoods which are most similar in the sample are selected to serve as synthesis pixels, pixels of a target image are synthesized according to the sequence of scanning lines, and after synthesis, color quantization is carried out on the image by means of the clustering algorithm to generate a final camouflage image. By means of the method, similar camouflage images can be synthesized quickly according to the background, and the application aim of quickly camouflaging a target in a self-adaptation mode can be achieved.

The present application provides a robust, illumination invariant apparatus and method for detecting and recognizing various traffic signs. A robust method for detecting and recognizing the traffic signs using images captured by a digital color and night vision camera, the said method characterized in being illumination invariant comprising the processor implemented steps of: transforming RGB image into HSV color model and subsequently extracting desired color components by using color quantization; filtering the noise components in the HSV color model based on object symmetrical shape property; detecting edges of the objects and subsequently detecting the distinct objects in the noise components filtered image; classifying the shapes of the traffic signs based on shape of the determined distinct objects; and recognizing the classified shapes of the traffic signs by template matching. Further, the method provides the provision for warning the driver by use of the recognized data of the traffic signs.

The invention discloses a video key frame extraction method based on color quantization and clusters. The method comprises the steps of loading video data flow; conducting single frame scanning on video flow; conducting the color quantization on obtained frame images, and extracting main color features of the frame images going through quantization; calculating similarity of adjacent frames so as to obtain adjacent frame difference; conducting shot boundary detection according to the adjacent frame difference; conducting shot classification on intersected shots and extracting a representative frame of each shot; and conducting compression clustering on the sequence of the representative frames so as to obtain a key frame sequence. According to the method, the color quantization is conducted on the single frame images so that main color of the images is extracted, frame difference calculation is conducted through the cluster feature similarity calculation method based on color features of the clusters so that the shot boundary detection is realized, and finally clustering according to the compression ratio is conducted on the extracted representative frames. Due to the fact that he whole process is low in dependency on video formats and types, the method has good universality and adaptability, is simple in calculation and low in space consumption, and can effectively avoid the phenomenon of key frame selection redundancy, control the number and quality of the key frames, and realize control of the video compression ratio.

A new color model HMMD is disclosed based upon hue, the shade, the tone, the tint and the brightness of a color, and a color quantizing method using the HMMD color space. The present invention regards the colors adjacent to a given color in the color space as the neighboring colors to consider indices of the neighboring colors as the partial values according to the distance. Accordingly, the present invention can accomplish a color quantization close to the change of the color sensed by the human eye, thereby capable of enhancing a performance of the image searching system based on content.< / PTEXT>

Automatic interpretation processing operations, such as character recognition, require a binary image of information-bearing image elements and a background. Digital image data produced by scanning a color-containing document frequently contain many different colors so that a division into information-bearing image elements and background is not evident. The invention divides connected components (contiguous pixels with the same color) in the digital image into background and other connected components, whereafter the other connected components are allocated, in accordance with a predetermined criterion, either to the background or to a foreground. The foreground connected components are combined into information elements suitable for the automatic interpretation processing. The division into information-bearing image elements and background can be preceded by a color quantisation processing of the digital image data.

The invention relates to a vector-quantization-based overall and local color image searching method. The invention provides a new color image searching method, and relates to the field of an image processing technology. The method comprises the steps of converting an RGB (red, green and Blue) color space into an HSV (Hue, Saturation, Value) color space, performing relatively accurate clustering division on the color space by applying a neural-network-based competitive learning algorithm training code book; describing a space distribution situation of colors by introducing a color transfer matrix; combining the two characteristics, namely an index column diagram and a main color transfer matrix, so as to perform the similarity measurement; processing images by applying a morphological opening-and-closing operation, highlighting a target contour so as to extract a local interest area so as to highlight an important area and limit background information. The color image searching method overcomes the defects that the color space distribution description is not enough and the background information cannot be effectively limited by using the overall color histogram method. By using the vector-quantization-based overall and local color image searching method, the color quantization is relatively accurate, the matching effect is relatively good, and the vector-quantization-based overall and local color image searching method is an effective method for further improving the researching efficiency.

The invention discloses a high-resolution remote sensing image segmentation method based on inter-scale mapping. The high-resolution remote sensing image segmentation method has the advantages that aiming at geographical object extraction involved in objet-level change detection, the high-resolution remote sensing image multi-scale segmentation method based on wavelet transform and an improved JSEG (joint systems engineering group) algorithm is provided; aiming at a key problem that a conventional JSEG algorithm affects segmentation accuracy in high-resolution remote sensing image segmentation, a corresponding improvement strategy is adopted to achieve a good effect; wavelet transform is introduced to serve as a multi-scale analysis tool, excessive rough color quantization of the conventional JSEG algorithm is abandoned, and thus, detail information in original images is kept to the utmost; a novel inter-scale segmentation result mapping mechanism is set up, image segmentation of a current scale is realized on the basis of a segmentation result of a previous scale, and correction of the segmentation result of the previous scale is also realized, and thus, accumulation of inter-scale segmentation errors is decreased effectively; finally, an improved multi-scale segmentation strategy and an improved region merging strategy are provided, and over-segmentation and mistaken merging are effectively reduced.

A color space structure and quantizing method is disclosed. Upon conversion from the RGB color space into HSV color space, a difference value S' (S'=max-min) is obtained by varying the maximum value of the saturation S in accordance with the value V resulting in a HS'V color space where the difference value S' forms one axis of the color coordinate, such that all of color areas are evenly distributed and the variation of color is uniformly displayed.

A method of compressing a color image is provided. The color image comprises color data for a plurality of pixels. The method includes the step of obtaining red, green and blue pixel values of an object of interest in the image. A calculation is made of the complement of the red, green and blue values of the object of interest. Transformation coefficients are calculated which transform the complements of red, green and blue values of the object of interest into representations in a transformation color space. The transformation coefficients are applied to all the pixels in the image to thereby obtain a transformed data set representing the image having components along three mutually orthogonal axes (A, B and C herein) in a three-dimensional transformed color space. The transformed data set is scaled in accordance with the color quantization used in the system; e.g., the A, B and C values are between 0 and 255 for an 8 bit quantization. A compression algorithm, e.g., a loss less algorithm such as WINZip or LZW is applied to at least two components of the transformed data set to thereby produce output data representing a compression of the image.

The invention discloses a method and device for making an image cartoony and fusing the image with a 2D code, and relates to the technical fields of image processing and 2D codes. The method comprises that an input image is made cartoony; a standard 2D code is generated according to an input text; and the cartoony image is fused with the standard 2D code to form a cartoon 2D code. When the image is made cartoony, a Gaussian difference operator is used to describe a contour line along the tangential flow field in the edge, and double-side Gaussian smoothing and color quantification are carried out. When the image is fused with the 2D code, fusion is carried out on the basis of a module, and under the condition that content included by the 2D code is not changed, the visual display effect of the 2D code approaches a binary result of the cartoony image as possible; and fusion is carried out on the basis of image pixels. The device comprises an image cartoonlization module, a standard 2D code generation module and a cartoony 2D code fusion module. According to the invention, image cartoonlization effectively remove concrete content of the image, so that the 2D code is fused with the image in a better way, and the fused 2D code is more beautiful.

A method and apparatus for performing color halftoning using block quantization is disclosed. In one embodiment, the method comprises dividing an input image into a plurality of blocks, clustering pixels in each of the plurality of blocks into a plurality of color groups, block quantizing each of the plurality of color groups for each of the plurality of blocks to a representative color value, and creating an output halftone image having a distinct block associated with each block in the plurality of blocks, with the representative color value for each group in each distinct block being approximated over a spatial area covered by the group in said each distinct block.

The invention relates to a fabric image color feature extraction method. According to the method, a color space quantification improvement method based on an essential color clustering method is provided during image color quantification grading, so that gray level quantification is better conducted on the color of each pixel in an image; meanwhile, the essential color histogram of the image and the color ratio information of each essential color are extracted with the essential color weight histogram, so that the accuracy of fabric image retrieval is improved and a fabric image meeting requirements of users is retrieved efficiently.

The present invention relates to a multimedia retrieval system and a method thereof which is capable of performing a multimedia data retrieval among different systems using color histograms constructed with different color spaces and color quantization methods, in particular to a content-based multimedia retrieval system and a method thereof which is capable of retrieving multimedia data among different systems, by extracting or converting a color histogram of a query image or a color histogram of an image to be retrieved into a color histogram of the same color space and color quantization method each other, and using the extracted or converted color histogram.

The invention discloses an outdoor traffic sign identification method under a low-illumination scene. The method comprises the following steps: allowing a low-illumination scene image containing traffic signs to be subjected to HSV conversion, calculating an integration histogram of the low-illumination scene image and detecting a traffic sign interest area through a threshold value method and a mathematical morphology method; adopting an SIFT algorithm to calculate descriptors of scale-changeless characteristic points of the traffic sign interest area and matching the descriptors with an image in a standard library so as to obtain matched point pairs; taking the coordinates of the scale-changeless characteristic points as image center spots to obtain subimages; respectively calculating the color-improved quantification histograms, which are fused with color tone, saturation and brightness, of the subimages, as well as the normalized color-improved histograms; calculating an included angle of two histograms of two matched points of the matched point pairs; removing the matched point pairs with big-different included angle; removing the matched point without a matching point; and obtaining the identification result of the traffic signs of the low-illumination scene image according to the number of the matched points. The method provided by the invention has the advantages of stronger robustness, and higher accuracy rate.

The invention discloses a refined embroidery texture migration method based on a sample domain, which comprises the following steps: in a model training stage, based on a cyclic generative adversarialnetwork, adding a Waserstein loss item, a feature matching loss item and an MS-in loss item into a loss function, and performing feature matching on the loss item and the MS-in loss item; according to the SSIM loss item, a self-attention mechanism is added to a generator G network, and a resolution enhancement method from coarse to fine is introduced. A training model is used for synthesizing embroidery textures; performing color quantization processing based on self-organizing network mapping on the input embroidery vector design drawing, performing color correction on the generated embroidery textures through the quantization method, and finally combining the embroidery textures with the generated normal map to obtain an embroidery texture map which meets the three-dimensional display requirement and has both embroidery colors and concave-convex textures. According to the method, rapid generation from image design to refined embroidery texture is realized, and the rework rate of production personnel is reduced while the communication efficiency of the design personnel is improved.

The invention relates to the technical field of computers, provides a page color adjustment method and device and electronic equipment, and aims to solve the problem of low adjustment efficiency caused by tedious operation in a color adjustment process. The method comprises the steps of conducting color quantization processing on a first image in a first page, acquiring N first candidate colors ina first color space wherein N is an integer larger than 1; converting the N first candidate colors into N second candidate colors in a second color space; determining a target color based on the N second candidate colors; and adjusting the color of a first region in the first page according to the target color. The invention can improve color adjustment efficiency.

The invention relates to an automatic generation method for an imitated camouflage pattern, and the method comprises the steps: obtaining a target image and a background region image; selecting a region for imitating the camouflage pattern; generating an imitated background image; carrying out the camouflage color quantification of the generated imitated background image and the extraction of a main color; carrying out the spot dividing of the generated imitated background image according to the extracted main color, and generating the camouflage pattern. According to the invention, the method achieves the precise and targeted imitation of the background region after the destroying of a target, and a high-precision background camouflage image which accords with the visual perception effect of human eyes can be generated, thereby guaranteeing the fusion effect of the generated camouflage image and the destruction area background, and improving the efficiency of a background imitation algorithm.

The invention discloses a color quantification method based on a density peak value, and the method comprises the steps: letting an image comprise N pixel points, calculating the Euclidean distances dij between each two pixel points, and building a distance matrix D; letting a cut-off distance be dc, and calculating a local density rho (i) of each pixel point according to the distance matrix D; for each pixel point, finding all pixel points which have higher local densities than the pixel point; finding the Euclidean distances from the pixel point to the pixel points with higher local densities according to the distance matrix D, and defining a minimum Euclidean distance delta (i); calculating a parameter gamma (i), equal to rho (i) * delta (i), of each pixel point, enabling all parameters gamma (i) to be arranged in a descending order, and selecting the pixel points corresponding to the front K parameters gamma (i) as cluster center points; enabling each pixel point after the K pixel points to be ranked in types corresponding to the cluster center points, and then employing a mean value of each type to represent the pixel value of this type, thereby completing the color quantification. The method is good in performance, can effectively and robustly achieve color quantification, and can flexibly meet different needs.

The invention discloses an intelligent color matching method for textile patterns. The method comprises the following steps of: (1) extracting a color matching scheme from popular color patterns, design patterns and digital photos with professional color matching levels, and establishing an expert color matching scheme database; (2) inputting patterns through color quantification and re-clustering extraction to acquire a group of color matching scheme capable of representing the input patterns; (3) respectively sequencing the representative colors in the input pattern color matching scheme and the expert color matching scheme according to the proportion of the representative colors in the patterns, and establishing mapping between the representative colors in the input pattern color matching scheme and the expert color matching scheme according to the sequencing result; and (4) performing representative color replacement on the input patterns according to the established mapping relation between the input pattern color matching scheme and the expert color matching scheme, and keeping the color level of the original input patterns. By the method, the professional level of the colormatching scheme is improved when a user customizes textile products.

A color quantization method in RGB color space which preserves high precision of luminance information in an original high precision RGB image signal which is quantized to a lower precision (lower bit depth) RGB signal. The method can be used to convert the original RGB signal to arbitrary quantization levels in RGB space.

A color quantization or re-quantization method is provided that combines two dimensional halftoning with luminance preserving quantization (LPQ) for better perception results of high precision color video quantization. A combination of LPQ and error diffusion, and a combination of LPQ and spatial dithering, is provided. To combine LPQ and spatial dithering, the spatial dithering is regarded as a two-step processing, a mapping and a simple rounding. To combine LPQ and dithering together, a rounding step is replaced by the LPQ algorithm in the combination. Further a method is provided for post-processing which is applicable to both cases to reduce the color perception for grayscale image.

The invention discloses a base color and index map algorithm for correcting inter-frame residual errors of a screen video. In the algorithm, the correlation of the inter-frame residual errors is discovered and utilized so as to acquire an efficient coding compression ratio. The residual errors are converted into base colors and an index map through color quantization, and the traditional code conversion and quantization process is omitted. In the color quantization process, the YUV combined color quantization method based on the K-Means algorithm is adopted. Then, the context-dependent hierarchical entropy coding method is adopted in entropy coding of the index map. Finally, the algorithm is applied to standard reference software for HEVC range extension as a new mode for correcting the inter-frame residual errors.

A method for assigning a color symbol to an image pixel comprises selecting a luminance value from a discrete set of quantized luminance values; selecting a chrominance value from an ordered discrete set of quantized chrominance values; and composing a color symbol from an index of the selected luminance value and an index of an ordinal position of the selected chrominance value. In particular embodiments of the invention, each discrete chrominance value is selected from a Fibonacci lattice on a constant-luminance plane in a perceptually uniform color space such as Lab or Luv.

A computer system (710) receives a desired upper bound δSmax for a relative quantization step ∥S′−S″∥ / ∥S′∥ to be used when quantizing any color in some range of colors. Here S′ and S″ are adjacent colors in the set of colors to be made available for the quantized image, and ∥·∥ is a norm in a 70%-orthonormal linear color coordinate system, the norm being the square root of the sum of squares of the tristimulus values. The computer system determines (510) suitable quantization steps for the brightness coordinate (B) and the chromatic coordinates (e,f) in a non-linear color coordinate system, and quantizes (520) the brightness and chromatic coordinates accordingly.

A method and system for describing image region based on color histogram is provided. The method for describing an image region based on color histogram comprising: a color quantization step for quantizing said image region into a quantized color region, wherein a specific color range and colors outside the specific color range are differently treated in the quantizing; a color histogram calculation step for computing one or more color histograms according to said quantized color region; and a histogram assembling step for generating a unique vector by using said one or more color histograms.

The invention discloses a machine vision-based method for conducting the color quantization of a colored bottle cap image. The method comprises the steps of (1) collecting the image of a colored bottle cap; (2) according to the type of the bottle cap, selecting the number K of colors required for the image after being processed; (3) according to the selected number K of colors, subjecting the collected image of the colored bottle cap to color clustering quantization; (4) according to a judgment condition for ending the clustering, judging whether the clustering is completed or not, and obtaining K color types after the completion of the clustering (clustering results are different for different kinds of bottle caps); (5) for the images of the colored bottle cap obtained later, assigning colors obtained during the previous clustering step to a new image according to the principle of shortest Euclidean distance. The method is applied to the pretreatment process for the defect detection of bottle caps. Based on the method, the basic color features of processed color images are retained, and the influence of light sources, lens and other environmental factors on the imaging quality is reduced.