SIFT-based multi-frame super resolution for 250 million pixel images
Katsuhisa Ogawa, Yuri Yamaguchi, Yutaro Iwamoto, Xian‐Hua Han, Yen‐Wei Chen · 2016
In this paper, we propose a SIFT-based multi-frame super resolution for 250 million pixel images. In the proposed method, we first use the SIFT operator to detect key points in each frame. Then we use a closest matching method to find the correspondence among multi-frame images. The corresponding key points are used to register multi-frame images to a reference image, which is randomly selected from the multi-frame images. After registration, we combine the aligned multi-frame images to form a high-quality and high-resolution image. We applied the proposed method to enhance the quality of 250 million pixel images, which is obtained by the Canon's 250Mpixel CMOS-image-sensor.