Detection and restoration system for mucous membrane image reflective area of endoscope

Abstract

The invention discloses a detection and restoration system for a mucous membrane image reflective area of an endoscope. The detection and restoration system comprises a data acquisition module, a mucous membrane reflective area detection network model module, a mucous membrane reflective area detection network training module, a mucous membrane reflective area detection module and a mucous membrane reflective area restoration module. The data acquisition module acquires a data training set, a verification set and a test image of an endoscope image, network model parameters are obtained through training and verification of the network training module, the detection module calculates the input endoscope image according to the network model parameters and marks and outputs a mucous membrane reflective area marking image, and the restoration module judges whether the endoscopic image has the mucous membrane reflective area or not, and repairs the endoscopic image with the mucous membrane reflective area to remove reflected light. The method can automatically adapt to endoscope reflective area detection; and after reflective endoscope image data is given, a repaired endoscope image can be quickly obtained, and thus the endoscope mucous membrane image reflective area detection and repair functions are achieved.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a system for detecting and repairing reflective areas of endoscopic mucosal images. Background technique [0002] Endoscopic equipment is mainly used to check the lesions of the human body orifices such as the digestive tract, trachea, ear, nose and throat. Doctors mainly judge the disease status by observing the video or image of the human cavity mucosa taken by the endoscope. Due to the illumination of the light-emitting device on the endoscope, the image usually has mucous membrane reflections. These strongly reflective areas block the mucous membranes, causing interference for doctors to observe and review the images. Therefore, the use of computer to automatically image the reflective area and remove or repair it will help facilitate the diagnosis of doctors. [0003] Endoscopic images are widely used. Detecting and repairing the light-emitting area of ​​the image...

AI Technical Summary

Problems solved by technology

At present, the processing technology of endoscopic image reflection is mainly through some methods of fixed mathematical models, or the method of setting threshold to filter out the reflection area and then repair it. This type of method is more complicated. How to design a simple given reflection endoscopic It is a problem to be solved in this field that the image data of the endoscope and the method of automatically adapting to the detection of the reflective area of ​​the endoscope

Images