Infrared target-based selective encryption by chaotic maps

The increasing transmission for multimedia has induced a growing requirement for security of crucial content. Towards this direction, several selective encryption schemes have been proposed in literature, whereas most of them only consider the edge information as significant information rather than regions of interest (infrared targets). In this study, we propose an infrared target encryption scheme based on block-cross encryption and Logistic-Sine system. Initially, the infrared target regions are efficiently detected by a geometric active contour model based on a partial differential equation (PDE) formulation. When the detection regions are irregular, an appropriate rectangular containing the detection regions should be designed. Next the pixels of extracted infrared targets are encrypted using block-cross encryption mode based on the Logistic-Sine system. Results demonstrate that the significant information of an infrared image has been largely hidden. The keys about the cipher image are extremely sensitive and the proposed scheme has high security.