A Ultrasound Liver Image Enhancement AlgorithmBased on Multi-Scale Retinex Theory

The paper proposes a B-ultrasound image enhancement algorithm based on Multi-Scale Retinex theory (MSR). Gaussian function of three-scale is adopted as surround function to process B-ultrasound liver cirrhosis images. When standard deviation σ is bigger, enhanced image color fidelity is better, and when σ is smaller, the image has better dynamic compress impact and details. In order to prove the algorithm´s effectiveness, quality evaluation such as lightness, contrast and entropy are extracted. The impacts of five kinds of methods including histogram equalization, wavelet image de-noising, homomorphic filtering, Single-Scale Retinex (SSR) algorithm and MSR are compared. Experimental results show that the method proposed based on MSR can effectively improve vision quality of image and assist doctor in clinical diagnosis.