Font Size: a A A

Research On Algorithms Of Improving Digital Photograph's Several Image Quality Problems

Posted on:2011-05-24Degree:MasterType:Thesis
Country:ChinaCandidate:J H ZhouFull Text:PDF
GTID:2178330338979784Subject:Instrument Science and Technology
Abstract/Summary:PDF Full Text Request
When we take digital photographs, due to illumination, weather and the camera parameters'settings, the image quality will be degraded. Digital photos can be improved by image processing methods which are different from film cameras.The paper analyzes several reasons of the bad-quality digital photographs and researches algorithms for improving kinds of photographs taken on different situations. This paper studies mainly in the following respects:Firstly, aimed at sharpening photos in the low-visibility conditions, the gray and color image enhancement algorithms are studied. The paper studies the theories of histogram equalization, homomorphic filtering and wavelet-based homomorphic filtering, then improves method of homomorphic filtering based on wavelets and applies it to HSI color space. According to the experiments of enhancing low-visibility photographs in various weather conditions, the conclusion is reached that the modified wavelet-based homomorphic filtering is effective for low-visibility photographs.Secondly, the paper researches the algorithms for correcting the unevenness of illumination in digital photographs Retinex algorithm is used and improved. The enhanced effect of the algorithm is better by clipping and drawing histogram adaptively. Finally, it is improved that the modified method of Retinex algorithm can eliminate uneven illumination in color digital photographs.Thirdly, the color offset is corrected with white balance algorithm. The paper analyzes several white balance methods such as Gray World Method, Perfect Reflector Method and an automatic white balance method based on edge detection. The experiments indicate that the third method achieves the best effect and is a widely used method.Finally, digital zoom is applied for low-resolution photographs. In the process of image super-resolution reconstruction, there is a need for interpolating pixels in the low-resolution image. The paper proposes a wavelet-based bi-cubic interpolation by a comparative analysis of the traditional interpolation algorithm. Experiments show the effectiveness of the proposed algorithm.
Keywords/Search Tags:digital photograph, unsuitable exposure, low visibility, white balance, digital zoom
PDF Full Text Request
Related items