In this thesis, I study high dynamic range imaging. My main work consists of a technique to lower the dynamic range to reproduce images with high dynamic range in low dynamic range and a technique toincrease the dynamic range to reproduce images with low dynamic range in high dynamic range display.<...
In this thesis, I study high dynamic range imaging. My main work consists of a technique to lower the dynamic range to reproduce images with high dynamic range in low dynamic range and a technique toincrease the dynamic range to reproduce images with low dynamic range in high dynamic range display.
First, when playing back an image with a high dynamic range in a low dynamic range display, if youignore data beyond the limit of the display through simple clipping, many visual information lost in the image is lost. Therefore, the technique of lowering the dynamic range (tone mapping) is very important, and a tone mapping technique considering the human visual system is required. For this purpose, I present a tone-mapping algorithm that minimizes the function that represents the visual sensation distortion that occurs after tone mapping. By exploiting human visual characteristics, I simplify the problem and find a closed-form solution that minimizes the function that represents distortion of visual sensation. In both subjective and objective evaluations, the proposed algorithm achieves a processed image that is most similar to the original image and has the best subjective image quality. On the contrary, when a low dynamic range image is reproduced in a high dynamic range display,the high dynamic range of the display is not utilized properly. Therefore, it is essential to extend the compressed low dynamic range image to the high dynamic range. For this purpose, I propose two approaches. First, by using the brightness adaptation model, I derive a brightness discrimination mapping that maximizes a function, which represents the local and global brightness discrimination range.Improvement of dynamic range is quantified by measuring increased discrimination ratio. In addition, since the information to be clipped in the image cannot be solved by the aforementioned method, It is necessary to further compensate the brightness by detecting the area (light source, etc.) to be further illuminated. Therefore, I propose the system to use the CNN to distinguish light sources from non-light sources, and use that information to identify regions that should be brightened.
In this thesis, I study high dynamic range imaging. My main work consists of a technique to lower the dynamic range to reproduce images with high dynamic range in low dynamic range and a technique toincrease the dynamic range to reproduce images with low dynamic range in high dynamic range display.
First, when playing back an image with a high dynamic range in a low dynamic range display, if youignore data beyond the limit of the display through simple clipping, many visual information lost in the image is lost. Therefore, the technique of lowering the dynamic range (tone mapping) is very important, and a tone mapping technique considering the human visual system is required. For this purpose, I present a tone-mapping algorithm that minimizes the function that represents the visual sensation distortion that occurs after tone mapping. By exploiting human visual characteristics, I simplify the problem and find a closed-form solution that minimizes the function that represents distortion of visual sensation. In both subjective and objective evaluations, the proposed algorithm achieves a processed image that is most similar to the original image and has the best subjective image quality. On the contrary, when a low dynamic range image is reproduced in a high dynamic range display,the high dynamic range of the display is not utilized properly. Therefore, it is essential to extend the compressed low dynamic range image to the high dynamic range. For this purpose, I propose two approaches. First, by using the brightness adaptation model, I derive a brightness discrimination mapping that maximizes a function, which represents the local and global brightness discrimination range.Improvement of dynamic range is quantified by measuring increased discrimination ratio. In addition, since the information to be clipped in the image cannot be solved by the aforementioned method, It is necessary to further compensate the brightness by detecting the area (light source, etc.) to be further illuminated. Therefore, I propose the system to use the CNN to distinguish light sources from non-light sources, and use that information to identify regions that should be brightened.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.