A Fast Image Mosaic Algorithm Based on Feature Partition Extraction
Zhiyou Lian, Jianhua Ren · 2024
This paper proposes a fast image stitching algorithm based on feature partitioning extraction to address the issues of long processing time, high computational complexity, and poor stitching performance in existing image stitching algorithms. This algorithm is implemented through two stages: optimizing image registration and image fusion. In the image registration stage, feature extraction and accumulation are carried out using image partitioning and an improved Gaussian pyramid layer series method; In the image fusion stage, an improved adaptive weighted average fusion algorithm is used for image fusion operations to improve stitching efficiency and make the image clearer and more natural. Experimental verification shows that the algorithm has strong robustness, significantly improving feature extraction speed and matching rate. Compared with mainstream algorithms, the stitching speed has increased by nearly 2 times, and the image related evaluation indicators have increased by about $5 \%$, basically meeting the real-time and timely needs of image stitching in daily life.