期刊名称:International Journal of Communication and Computer Technologies
印刷版ISSN:2278-9723
出版年度:2012
卷号:1
期号:2
出版社:IJCCTS
摘要:Human eye is perceptually more sensitive to certain colors andintensities and objects with such features are considered moresalient. Detection of Salient image regions is useful inapplications such as object based image retrieval, adaptivecontent delivery, adaptive region-of interest based imagecompression, and smart image resizing .This problem can behandled by mapping the pixels into various feature spaces. Thispaper proposed a methodology that encapsulate K-MeansClustering with Saliency map detection technique to determinesalient region in images using low-level features of luminanceand color and then extract the features. Proposed methodology issimple to implement, computationally efficient and generateshigh quality saliency maps and saliency object of the same sizeand resolution as the input image. Here presented schemeidentify salient regions as those regions of an image that arevisually more conspicuous by virtue of their contrast withrespect to surrounding regions.