Revate S.S.,Shri Chhatrapati Shivaji college |
Revate S.S.,Dr Bamuniversity |
Deshmukh P.D.,P.A. College |
Deshmukh P.D.,Dr Bamuniversity |
Kale K.V.,Dr. Babasaheb Ambedkar Marathwada University
International Journal of Applied Engineering Research | Year: 2015
In this paper, we have emphasis on multi fingerprint analysis for gender classification. As a growing demand in security and forensic investigation aspect day by day new solutions are designed to solve the issues. So, we have used 2500 samples of classified fingerprint database. 2-D Discrete Stationary Wavelet Transform (SWT) and Singular Value Decomposition (SVD) technique have used as a feature extraction methods. The fusions of these two extracted features are used in further process. At last, the Support Vector Machine Classifier is used to classify the datasets. The proposed fusion approach gives better results using SWT (Level 2) and SVD decomposition features. Thus, we have achieved 78% classification rate for females Right Index (RI) fingers and 75% for males Right Ring (RR) fingers. © Research India Publications.
Asif A.M.A.M.,I.M.S.I.T. |
Hannan S.A.,Vivekanand College |
Manza R.R.,Dr Bamuniversity |
2010 IEEE International Conference on Computational Intelligence and Computing Research, ICCIC 2010 | Year: 2010
In this paper, we propose the comparison of original text in the form of bitmap image and extracted image by using various fonts of text as a bitmap image. The purpose of data hiding is to embed relating textural description into an image. The textural description and original host image can be extracted and reconstructed from the stego-image in the data extraction process. However, the reconstructed host image will more or less be damaged or distorted by using traditional data hiding methods. The advantage of applying LSB data hiding is its simplicity in implementation and its robustness. However, it is not suitable for the data hiding of binary images because the original host image cannot be completely reconstructed due to the problem of information loss. Here we designed a program in Matlab using image processing toolbox which can encode the text images in a single color image. We can recover the image very easily. © 2010 IEEE.