Wireless capsule endoscopy images enhancement via adaptive contrast diffusion

Wireless capsule endoscopy (WCE) has been widely applied to diagnose diseases in human digestive tract due to its advantage that it can directly view the entire small intestine for the first time. However, many WCE images are rather dark, which challenge to analysis and diagnosis exerted by a clinician. To overcome this shortcoming so as to assist physicians, especially computer aided detection, we propose an adaptive contrast diffusion to enhance WCE images. Based on local analysis of WCE images, we put forward a new idea of contrast diffusion. Then, we employ contrast diffusion to enhance WCE images with an adaptive choice of the conductance parameter, which plays an important role in diffusion. Extensive experiments demonstrate that this new method exhibits promising performance of enhancement for WCE images, leading into a better visualization as well as an improved classification performance of WCE images using computerized methods.

► With local analysis of a WCE image, we propose a new idea of contrast diffusion.
► We utilize contrast diffusion to enhance WCE images.
► An adaptive choice of the conductance parameter in the diffusion is also proposed.
► Experiments show an encouraging performance of enhancement for WCE images.