AUTOMATIC BAND SEGMENTATION AND DETECTION OF 1D-GEL IMAGES WITH DIFFERENT EXPOSURES

Abstract

Many gel electrophoresis image segmentation methods have been proposed. However, most of the proposed methods cannot provide a satisfying performance since a gel electrophoresis image is a manually captured image often with various noise and poor quality. This paper presents a one-dimensional-gel image segmentation method (1DGISM) to cut the bands off from a one-dimensional gel electrophoresis image (1D-gel image). The experimental results showed that 1DGISM can give an impressive segmentation performance even if the band boundaries are very indistinct. Moreover, the authors suggest that a low exposure for imaging be helpful in segmenting bright bands. Alternatively, taking the image at a high exposure is useful if the bands are dark.