Robust specular reflection removal and visibility enhancement of endoscopic images using 3-channel thresholding technique and image inpainting

Wooju Lim

doi:10.47577/technium.v2i7.2164

Abstract

Specular reflections create artifacts in endoscopic images, which may lead to misdiagnosis. In this paper, we propose a method for robust removal of specular reflections by using a thresholding technique in each of the RGB channels to segment the specular reflections from images. We further use dilation to ensure full local segmentation and inpainting to replace the areas of reflections with non-specular regions. Our method also provides a visibility enhancement feature to improve the decreased brightness due to the reflection removal by using the gamma-correction, histogram shift, and histogram equalization. On the Iparkmall Clinic dataset, our method has achieved average Peak-to-Signal-Noise Ratio (PSNR) of 42.62dB with a standard deviation of 5.80 dB and a minimum value of 23.3 dB. The average processing time was 219ms, enabling average 4 5 frames per second (FPS) processing speed on an Intel i7 processor.

Highlights

Endoscopy is a gold-standard imaging modality for gastrointestinal cancer screening
Computer- aided approaches suffer from specular reflections in the endoscopic images Fig. 1, which arise from light reflecting from the wet surface of the mucous membrane [1]
Our results indicate that the proposed method located the specular reflections and removed them successfully

Summary

INTRODUCTION

Endoscopy is a gold-standard imaging modality for gastrointestinal cancer screening. During an endoscopy, a flexible tube with a camera and a light source is inserted into the patient’s body, which enables the clinician to visually analyze the internal organs. Computer- aided approaches suffer from specular reflections in the endoscopic images Fig. 1, which arise from light reflecting from the wet surface of the mucous membrane [1]. Flat polyps are often unrecognized or have their size miscalculated due to the specular reflections, despite the potential for development into a malignancy. This phenomenon leads to a lower chance of early-stage diagnosis and treatment of cancer [2]. There- fore, there is a clear need for suppressing specular reflections in order to enable accurate human and computer- aided diagnosis. We propose an adaptive method of threshold filtering to segment the specular re- flections and remove them by image inpainting

RELATED WORK

QUANTITATIVE ANALYSIS

DISCUSSION AND CONCLUSIONS