Performance Comparison of Row, Column, Full Slant Transform and PCA for Face Recognition

In this paper we propose fast face recognition system based on the 1-D Discrete Slant Transform (ST) row feature vector-RV and column feature vector-CV. This scheme is less complicated and needs less time as compared to ST of full image. It is observed that in this method for 95% image energy the coefficient requirement reduces drastically compared to PCA and full ST. Thus computational burden decreases and recognition time reduces. We have used standard ORL database and results obtained are accurate. The algorithm is also tested for locally created database of male and female faces which gives accuracy around 90%. KeywordsSlant Transform (ST), Eigenfaces , Energy compaction, Face recognition ,PCA, Row feature vector, Column Feature Vector.