Study Of Image Enhancement And Image Encryption Algorithm Used For PACS System

Nowadays, digital image processing has been widely used in many fieldssuch as the military field, remote sensing, public security, and the biomedicalarea. Image enhancement and image encryption are two significant digital imageprocessing technologies. Image enhancement can enhance the details of images,while the image encryption technology can make images visually unidentifiable,so as to hide the real information in the images.With the support of PACS (Picture Archiving and Communication Systems),Digital Hospital has already become the development trend of modern medicalcare. As two fundamental steps in PACS, Image enhancement technique is usedto enhance image definition, so as to make diagnosing easier, while imageencryption technology can protect users’ privacy, so that image data can be stored,transmitted and used more safely. The studies in this paper have very importanttheoretical and practical values in promoting the development of the medicalimage enhancement and encryption technologies.Image enhancement in PACS mainly uses image enhancement algorithmswhich are based on the Retinex method and mathematical morphology. TheRetinex method is introduced to obtain a constant gain of the image. However,the Retinex method makes the image excessively enhanced in the edge regionswhich is of a relatively high contrast ratio and in relatively smooth regions,therefore halo may appear in the edge regions, and smooth regions may be over-enhanced. To solve this problem, in this paper a new PACS image enhancementalgorithm which is based on Retinex theory is proposed. In the new algorithm,the enhanced images are achieved through global and local image processingrespectively. On the other hand, in image processing nonlinear diffusionmodels are introduced to form the P-M model, so that de-noising effect can be obvious, but texture details and other features may be lost with this model. Inorder to solve this problem, in this paper a new PACS image enhancementalgorithm based on P-M model is proposed. The algorithm is deeply analyzed inrespect of multiple indicators such as brightness, definition, contrast, entropy,absolute average square error, peak average square error, normalized averagesquare error, peak signal-to-noise ratio, absolute average error, structuralsimilarity and average gradient.Image encryption algorithms in PACS system mainly uses imageencryption algorithms which are based on chaos theory. On the basis ofexisting methods, this paper has conducted researches on PACS imageencryption algorithms focusing on improving security and speed of imageencryption and decryption. A fast bit-plane based PACS image encryptionalgorithm is proposed, which is based on initial condition sensitivity of chaostheory. The proposed algorithm encrypts images in bit planes while guaranteeingthe encryption security. A large number of experiments including effect analysis,speed tests, and anti-attack ability tests have been carried out on many differentkinds of images. For smaller images, and when security is highly required, afinite-field based PACS image encryption method is proposed in this paper. Bycombining chaos mapping and finite field theories, the proposed algorithm cangreatly improve security. The experiment results demonstrate that the algorithmcan be applied to images and environments where security is strictly required.