Résumé:
Handwritten digits recognition is a key research problem in the domain of image analysis and pattern recognition. Specifically, the appearance approaches based on feature extraction to solve many research issues. This work presents a novel way to use encryption techniques such as feature extraction stage. Moreover, this method is very efficient for handwritten digit recognition as it is less effected from variations in shape and slant. The proposed method is carried in two steps: first, we concatenate entire image with its encrypted image. Second, the pixels imageis collected into a vector. The performance of the proposed approach is evaluated for recognizing the isolated handwritten digits from the CVL dataset. The experimental results showed recognition with high accuracy.
