According to the requirement of real-time image processing and the variety of noise, a hybrid filter algorithm is presented in this paper, and then the color image real-time denoising is realized on FPGA. An improved median filter algorithm which combines the number of roots with the comparison of ranks is presented upon the conditional median filter algorithm. Based on the local threshold, the hybrid algorithm which combine improved median filter with weighted mean filter judges the noise type, and then uses different denoising method. At last, the algorithms are finally fulfilled in the real-time image acquisition system. The results show that the hybrid filter algorithm satisfies the real-time requirement and have higher visual effect. © 2011 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Liang, F., & Lin, Y. (2011). Low illumination image denoising research and its real-time implement. Advances in Intelligent and Soft Computing, 105, 165–170. https://doi.org/10.1007/978-3-642-23756-0_27
Mendeley helps you to discover research relevant for your work.