Endoscope image mosaic based on pyramid ORB

• Published in: Biomedical signal processing and control Vol. 71; p. 103261
• Main Authors: Zhang, Ziyan, Wang, Lixiao, Zheng, Wenfeng, Yin, Lirong, Hu, Rongrong, Yang, Bo
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.01.2022
• Subjects: Endoscope; Image fusion; Image mosaic; Image registration; pyramid ORB; Image registration; Image mosaic; pyramid ORB; Image fusion; Endoscope
• ISSN: 1746-8094; 1746-8108
• DOI: 10.1016/j.bspc.2021.103261

•The key to endoscopic image mosaics success is the accuracy of image registration and fusion.•This paper uses the Gaussian Pyramid to improve the simple ORB-oriented algorithm.•The improved method reduces the calculation and storage requirements and improves the image registration efficiency.•The pyramid ORB algorithm solves the problem that the ORB does not have scale invariance.•The pyramid ORB is more suitable for minimally invasive surgery endoscopic image mosaic. The primary method of endoscopic image mosaics seamlessly mosaics several continuous and overlapping endoscopic images to have a broader vision and clearer images. The key to its success is the accuracy of image registration and fusion. This paper uses the Gaussian Pyramid to improve the simple ORB-oriented algorithm. The experimental results show that the pyramid sphere algorithm has invariability, good robustness in scale change and rotation change, high registration accuracy, and the stitching speed is about 10 times that of SIFT. The improved method has more efficient feature extraction and expression methods, reduces the calculation and storage requirements, and improves the image registration efficiency. The pyramid ORB algorithm solves the problem that the ORB does not have scale invariance. The pyramid ORB is more suitable for minimally invasive surgery endoscopic image mosaic through theoretical analysis and experimental verification. The algorithm has good robustness and real-time performance.