There are many active research projects accessing and applying shared ADNI data. Use the search above to find specific research focuses on the active ADNI investigations. This information is requested annually as a requirement for data access.
| Principal Investigator | |
| Principal Investigator's Name: | fish little |
| Institution: | Xi‘an University of Science and Technology |
| Department: | College of Mechanical Engineering |
| Country: | |
| Proposed Analysis: | Images captured in low light conditions suffer from unfavorable visualization, low contrast, color distortion, and noise, which makes those images were unsuitable for computer vision tasks. While it was inevitable that tackle computer vision mission on low light images. This paper proposed an image enhancement method based on HSV space. The basic idea was that the Value component decided the illumination of an image, and the Saturation component contains more Gaussian noise. We followed the idea of divide and rule, denoising on the Saturation and increasing illumination on the Value. Then applied the Bayesian rule to fuse the Saturation and Value and then merged the three components to RGB format. Finally, a Semi-implicit based ROF model was introduced to denoise global noise to obtain the enhanced image. Such an integrated method allows us to enhance image more clearly. Extensive experiments on the LOL dataset shown that the proposed method is very competitive and improves the image quality significantly. |
| Additional Investigators | |
| Investigator's Name: | fish little |
| Proposed Analysis: | Images captured in low light conditions suffer from unfavorable visualization, low contrast, color distortion, and noise, which makes those images were unsuitable for computer vision tasks. While it was inevitable that tackle computer vision mission on low light images. This paper proposed an image enhancement method based on HSV space. The basic idea was that the Value component decided the illumination of an image, and the Saturation component contains more Gaussian noise. We followed the idea of divide and rule, denoising on the Saturation and increasing illumination on the Value. Then applied the Bayesian rule to fuse the Saturation and Value and then merged the three components to RGB format. Finally, a Semi-implicit based ROF model was introduced to denoise global noise to obtain the enhanced image. Such an integrated method allows us to enhance image more clearly. Extensive experiments on the LOL dataset shown that the proposed method is very competitive and improves the image quality significantly. |
