Performance of Assorted similarity Measures in Gray Image Colorization using Thepade’s Cosine Error Vector Rotation Algorithm

The target gray scale image is colorized by procuring (extracting) color pixel values from relatively similar source color image training set. The proposed algorithm colorizes the target gray scale image using Thepade’s Cosine Error Vector Rotation (TCEVR) algorithm. The training set of the proposed algorithm consist of 25 vivid images. Distinct families of similarity measures are considered and from which selected similarity measures are opted for execution evaluation. Established on the ramifications (result) analysis from the implementation performed on different similarity measures where Chebychev, Manhattan preceded by Euclidean gives better performance. Although the substandard colorization quality is given by Hamming preceded by Jaccard.
This paper was published in IJICAR - International Journal of Integrated Computer Applications & Research - http://ijicar.com