Review: A Comparative Study based on Video Stitching Methods

• Published in: 2021 Second International Conference on Electronics and Sustainable Communication Systems (ICESC) pp. 1157 - 1163
• Main Authors: Jagadeeswari, M, Manikandababu, C S, Dhviya, M Sree, Meenakshi, J Varshini
• Format: Conference Proceeding
• Language: English
• Published: IEEE 04.08.2021
• Subjects: Communication systems; Computer vision; Dynamics; Errors in video stitching; Feature based method; Feature extraction; Heuristic algorithms; image stitching; Object detection; Object tracking; Position based; Stitching techniques; Video stitching
• DOI: 10.1109/ICESC51422.2021.9532635

Image/video stitching technology has been used in several fields like image and video processing, medical imaging applications, computer vision etc.., There are several image stitching methods and algorithms of which some are most popular and commonly used. Image stitching is the basis for video stitching. It can be easier to stitch images with a static background. Video stitching with static background is challenging but with a combination of various algorithms and techniques, it can be achieved. However, video stitching with dynamic elements is slightly harder as it results in errors like ghosting, blurring especially in the overlapping region between the images. This paper presents a deep survey based on the comparison of image stitching methods like feature based, position based and depth-based along with the algorithm in video stitching. It also presents some algorithms and concerns about the errors that arise during the image registering and stitching process. This paper also summarizes the advantages and limitations of various image and video stitching techniques and algorithms.