Main Article Content
Abstract
Image quantization is a process which helps in keeping image in computer with less space by depleting the count of colors in it so that it can be stored in computer using less space and Image is quantized such that perceived difference between original and quantized image is as small as possible. Basically Image Quantization is used in devices where storage requirement is small and it is used for displaying images in devices which support a few number of colors and also used to reduce the transmission bandwidth in multimedia applications. Image quantization has been implemented using different algorithms, mainly clustering and splitting algorithms has been used. In Splitting algorithms, by using some criteria the color area of original image is divided into two decouple cells. After that Unless the required number of cells not acquired, the splitting method is repeated. Then, the choice of cluster heads of the cells is made to be the illustrative colours in a color map and in Clustering algorithms we assign colors into group so that colors with similar properties are placed in one group and can be differentiated from the colors of other group. Two types of clustering algorithms are used which include pre-clustering and post clustering. Commonly pre-clustering approach is used because it do not require complex interrelationship between color clusters. Different techniques have been used to implement Image Quantization using these algorithms. In this paper these algorithms has been discussed and the performance of Image Quantization techniques have been compared on the basis of these algorithms.