Human Detection And Encryption Technologies Still Image

The implementation of human detection and encryption by computer is an indispensable technology in modern society. This technology has been widely used in the field of privacy protection, intelligent monitoring, public security, video games and soon.Human detection is achieved based on features and classifier. The human body is of many characteristics, such as human gesture, clothing, body, environment and environmental light intensity, which are various. No matter what characteristic is chosen to achieve accurate human detection is always difficult, because characteristics that are used in human detection must be stable and of little changes. Histogram of Oriented Gradient (HOG) feature is employed to achieve detection of standing human body in pictures. The detection consists of the following steps:firstly, a large number of positive and negative samples are collected and their HOG features are extracted. Secondly, these features are input to support vector machine (SVM) for training, obtaining a classifier can be used for detection. Finally, inputting images to be detected to this classifier performs detection. Experimental results show that the human body in images can be accurately detected using the above method.Next, detected parts of the body are encrypted for individual privacy protection. In exceptional circumstances, those encrypted parts can be decrypted if necessary. The encryption method used in this paper is based on fractional Fourier transform. Fractional Fourier transform is the extended form of traditional Fourier transform, and is appropriate for processing non-stationary signals, which not only covers shortages of Fourier transform, but also adds a parameter transforming order so that the security of image encryption is also enhanced.The extracted images are encrypted in this paper using fractional Fourier transform with certain order. That is to say, rotating images in any degrees enable people not to see original images. However, the usage of fractional Fourier transform with negative order can restore original images. Experimental results show that the encrypting results by fractional Fourier transform with distinct order are very different, and original images can be accurately restored by fractional Fourier transform with right order.