A selective bitplane image encryption scheme using chaotic maps

Abstract

Partial encryption is one of the viable solutions for low power, high speed, real time secure multimedia communication. In this paper, a chaotic tent map based selective bitplane encryption technique is proposed for both gray scale and color images. After decomposing the original image into eight bitplanes, each bitplane is classified into either significant or non-significant category by defining a flexible threshold value of 0.3, deduced experimentally. Following this segregation, only the significant bitplanes are encrypted with the key stream sequences generated by a chaos-based pseudo-random binary number generator. The cipher image is then transmitted through public channel. The proposed scheme has three important contributions v.i.z. a) determination of significant bitplanes, b) encryption of only the significant bitplanes leading to reduction in computational complexity and c) elimination of the need for separate channel for transmitting the information about the significant bitplanes. It is shown that the proposed partial encryption scheme saves around 35% computation on the image database used here. Different types of attacks against this scheme are also analysed to show the robustness of this approach.