Digital Image splicing Detection Using RIC-LBP Feature Extraction Technique

In this paper, we proposed rotation invariant co-occurrence among adjacent local binary pattern (RIC-LBP) based feature extraction technique for forgery detection. We use Standard Deviation filter (STD) to highlights the image pixel variation, RIC-LBP operator for feature extraction, and Logistic Regression Classifiers (LRC) for forgery detection to know the internal statistics of the image. LRC is a machine learning technique so directly used as a classifier on the entire data set. So it differs from SVM classifier. In this proposed work, we used two datasets, Columbia and DSO-1, to evaluate our proposed work. It gives better results compare to various state of the art.