Multi-Feature Fusion Algorithm Based on AR-Based Panoramic Image Processing Technology Inside a Wind Turbine
Abstract
Image fusion is not only a simple image synthesis and superposition process, it is also an organic combination of the advantages of the information contained in each image data. It can also improve the clarity and understanding of the image, achieve further image segmentation, target detection and recognition, and provide a better understanding of image processing and more effective information. This article introduces the multi-feature fusion algorithm based on AR-based panoramic image processing technology of the fan. This paper adopts a matching algorithm based on additional landmarks through the calibration of the vision system, completes the positioning of the landmarks in the space, so that the positioning of the landmarks can be tracked in real time in the process, and through the augmented reality image fusion method, a virtual 3D model is created using 3D software, positioning the fiducial point of the real world to be enhanced, and then fusing the real scene image and the virtual image graphics according to a certain relative position space. The experiment results show that the execution efficiency of the multi-feature fusion algorithm is 1.23 times better than the SIFT algorithm and its matching accuracy is 19.85% higher than that of the SIFT algorithm. This table reflects the performance of the multi-feature fusion algorithm for larger viewing angle transformation and contrast. The object image has strong robustness and matching performance in complex
situations.
Keywords: AR Technology; Fan Internal Structure; Panoramic Image Processing Technology; Multi-Feature Fusion Algorithm; Matching Algorithm
Cite As
J. Yu, J. Zhang, L. Ban, G. Feng, "Multi-Feature Fusion Algorithm Based on AR-Based Panoramic Image Processing
Technology Inside a Wind Turbine", vol. 30 no. 5, pp. 377-385, 2022.