APPLICATION OF OBJECT DETECTION AND IMAGE GENERATION IN CAPSULE ENDOSCOPY ASSISTED DIAGNOSIS SYSTEM

Abstract

The paper consists of a total of seven chapters. Chapter 1 introduces the background, aims and objectives of the research. Chapter 2 explores object detection models based on deep learning technologies. Chapter 3 introduces an invertible normalization module called InvNorm which can enhance the generalization ability of the model. Chapter 4 explores image generation technology to obtain much more capsule endoscopy images for research. Chapter 5 introduces the outline of the Capsule Endoscopy Assisted Diagnosis System and the details of each functional module. Chapter 6 introduces the patent of image generation, and shows the flow chart and result of the patented technology. Chapter 7 concludes previous chapters and discusses the limitations of this work. What is more, it also looks forward to the future development of this research.