Enhancement and colorization of infrared and other medical images

In this thesis, we study some image enhancement and colorization methods for medical and infrared images. First we summarize the characteristics of infrared images and survey different approaches of image enhancement and image colorization. Our objective is to develop a method for medical images, especially infrared images, which helps enhance the visual feature for medical diagnosis. We studied human visual characteristics-based image enhancement and intensity-based colorizing methods. We used wavelet transform with the human visual system considered, to enhance image contrast. The wavelet transform is applied to decompose the luminance information into approximate components and detail components. Then, a contrast enhancement technique for images based on the human visual system is applied to enhance approximate component and an inverse wavelet transform is applied so that contrast enhanced color image is obtained. For the purpose of making medical images have more visual appeal and provide medical personnel with more useful information. We study two colorization methods: the first one is based on a simple premise that neighboring pixels sharing similar intensities should have similar color and it requires human interaction, while the other one can automatically transfer chromatic information from a reference color image to the target image using pixel matching.