Please use this identifier to cite or link to this item: http://hdl.handle.net/10603/13433
Title: A probabilistic adaptive color reduction algorithm for single multiple images with without transparency
Researcher: Periasamy P S
Guide(s): Duraiswamy, K.
Keywords: Probabilistic Adaptive Color Reduction Algorithm (PACRA), Histogram Accumulated Matrix, Palette Accumulated Matrix, K-Means color reduction algorithms, Octree
Upload Date: 28-Nov-2013
University: Anna University
Completed Date: 
Abstract: Color reduction is essential to represent full color Red Green Blue (RGB) images where each pixel is described by three 8-bit color samples, in an approximate fashion by relatively small number of colors. The image display devices restrict the number of colors to be displayed simultaneously on a screen. One of the basic tasks of image processing based on Real Time Operating Systems (RTOS) and Digital Signal Processors is to reduce the number of colors in an image without or with minimal visual distortion. The methods described in the literature do not address the creation of a common palette for multiple images for simultaneous display, supporting images with transparency information and flexibility of adding colors by the user. To incorporate the above information in color palette creation, a new Probabilistic Adaptive Color Reduction Algorithm (PACRA) is proposed. Then all the colors in the palettes are accumulated into a single matrix. This matrix is called Palette Accumulated Matrix (PAM). The probabilities of colors in the PAM are accumulated into a single matrix and this matrix is called Histogram Accumulated Matrix (HAM). The PAM and HAM form the common data structure. Based on the experimental results of PACRA, the comparisons are made with Popularity, Octree, Median-cut and K-Means color reduction algorithms for a set of 14 natural images. The simulation result shows that, the PACRA improves the quality of the color reduced image remarkably in terms of both visual quality and quantitative metric. Mean Structural Similarity Index Measure (MSSIM) is used as a quantitative metric. In this research, a new image-dependent color reduction algorithm, namely, Probabilistic Adaptive Color Reduction Algorithm for single/multiple images with or without transparency information has been developed. The potential real-world applications of the algorithm are explored in detail. This algorithm can be easily implemented in different hardware devices. newline newline newline
Pagination: xviii, 143
URI: http://hdl.handle.net/10603/13433
Appears in Departments:Faculty of Information and Communication Engineering

Files in This Item:
File Description SizeFormat 
01_title.pdfAttached File17.79 kBAdobe PDFView/Open
02_certificates.pdf138.42 kBAdobe PDFView/Open
03_abstract.pdf24.48 kBAdobe PDFView/Open
04_acknowledgement.pdf13.4 kBAdobe PDFView/Open
05_contents.pdf74.21 kBAdobe PDFView/Open
06_chapter 1.pdf369.67 kBAdobe PDFView/Open
07_chapter 2.pdf305.06 kBAdobe PDFView/Open
08_chapter 3.pdf2.65 MBAdobe PDFView/Open
09_chapter 4.pdf646.22 kBAdobe PDFView/Open
10_chapter 5.pdf651.86 kBAdobe PDFView/Open
11_chapter 6.pdf17.66 kBAdobe PDFView/Open
12_references.pdf124.66 kBAdobe PDFView/Open
13_publications.pdf44.54 kBAdobe PDFView/Open
14_vitae.pdf36.97 kBAdobe PDFView/Open


Items in Shodhganga are protected by copyright, with all rights reserved, unless otherwise indicated.