| Summary: | 碩士 === 中原大學 === 通訊工程碩士學位學程 === 103 === Contrast is the key point of visual effects. Generally speaking, a greater contrast gives conspicuous images with plentiful color, while a lower contrast results in gray like images. The brightness and contrast are highly dependent. In general, an image with uniform brightness distribution would show a great deal of gray level detail and high dynamic range.
Currently on the market and the Internet, many image editing software packages provide automatic image enhancement functions for public use. With this kind of software, just one simple operation can improve the visual quality of the image. However, after a real test we found that, under certain conditions or special circumstances, part of the image enhancement functions does not work effectively, and there is a great chance of causing color problems. In other words, even after an image has been enhanced in brightness, it often induces the problems in hue shifting and poor saturation.
Therefore, the idea of the proposed method is to control the change in hue to a minimum and avoid the change of color attributes. We use histogram equalization to increase the dynamic range of luminance and render more options for saturation improvement based on the luminance change using our proposed technique called two-factor reconstruction. Because the psychovisual sense of color cannot be quantified, the proposed method provides an adjustable parameter for users to meet their color satisfaction. For convenience, the input image will be classified into several categories, and the parameter setting guideline for the category of that image will be provided to the users so that they can adjust the parameter to achieve the desired saturation of the output image.
The experimental results show that the proposed method successfully improves the performance of image brightness and contrast under the condition of preserving hue information, and simultaneously improve human visual experience by enhancing the saturation. It has a relatively good performance compared with others methods, and the statistics on psychological assessment also shows that using the parameter to adjust the image saturation can fully meet the needs of the users.
|
|---|
