A Novel Region of Interest for Selective Color Image Encryption Technique based on New Combination between GLCM Texture Features

摘要

Selective image encryption is an efficient way to reduce the amount of encrypted data that can achieve an acceptable level of security. Determining and choosing the region of interest in digital color images is a challenging task in selective image encryption due to their complex structure and distinct regions of varying importance based on instinctive feelings and opinions. Gray Level Co-occurrence Matrix (GLCM) is the core primitive method for texture analysis. To develop a novel selective encryption strategy, new features in acquiring and selecting Region of Interest (ROI) for the color images based on GLCM and Faster R-CNN for object detections is proposed. The Faster R-CNN is implemented for determining the boundary of an object in an image and the roughness criteria that representing the busy area are proposed for each object. The roughness state is based on developing new combinational relation between features for determining image regions that are developed upon the object texture analyses map blocks. The roughness block for each region is encrypted using AES Algorithm with a dynamic secret key generation based on sine and tan chaotic map methods and permutation of other blocks in each region. The security performance of selective image encryption is found to enhance considerably based on the rates of selective encrypted area. Thus, the proposed strategy achieves good alternatives, fulfills the desired confidentiality, and safe the privacy of the image.