A novel approach for image forgery detection using improved crow search algorithm

Abstract In recent days, the image data plays a important role in every field. In this digital world generation of image data is becoming popular. A novel method has been proposed for a forgery detection technique in image using enhanced improved crow search algorithm. Preprocessing is done at the initial stage to convert RGB to LAB space conversion, Proceeded with feature selection and alalysis using Ada boost algorithm and cascade architecture. Next step is to extract the features like Haralick feature, Skewness and inverse difference movement. In Continuation with feature extraction, feature selection is performed using proposed algorithm followed by enhanced convolution neural network classification. Finally the comparison of the proposed classification algorithm and sensitivity analysis is done.