Image Stitching using AKAZE Features
- PDF / 3,311,445 Bytes
- 13 Pages / 595.276 x 790.866 pts Page_size
- 51 Downloads / 333 Views
RESEARCH ARTICLE
Image Stitching using AKAZE Features Surendra Kumar Sharma1,2
•
Kamal Jain2
Received: 7 December 2019 / Accepted: 27 August 2020 Indian Society of Remote Sensing 2020
Abstract Accelerated KAZE (AKAZE) is a multi-scale 2D feature detection and description algorithm in nonlinear scale spaces proposed recently. This paper presents an image stitching algorithm which uses a feature detection and description algorithm; AKAZE and an image blending algorithm; weighted average blending. The whole process is divided into the following steps: First of all, detect feature points in the image and then get feature descriptors of detected points using AKAZE. Next, obtain corresponding matching pairs by using K-NN (K nearest neighbors) algorithm and remove the false matched points by MSAC (M-estimator SAmple Consensus) algorithm. MSAC is a variant of the RANSAC (Random Sample Consensus) algorithm and more accurate than RANSAC. Thereafter, calculate the homography matrix from correct matches. At last, blend the images by using weighted average blending algorithm. Comparison of proposed AKAZE-based algorithm with SIFT-, SURF- and ORB-based algorithms is also presented. According to the experiments, the proposed AKAZE-based image stitching algorithm minimizes stitching seam and generates a perfect stitched image, and also this algorithm is faster than previous algorithms. Keywords Image stitching AKAZE Image quality Panoramic image
Introduction Panoramic images or panoramas are wide-angle images that can produce a horizontal field of view up to 360. Panoramas can be obtained by stitching multiple images together using image stitching algorithm. It is a wellknown method of constructively extending field view of the camera. Panoramic imaging softwares implement image stitching algorithm to create large panoramas. Several panorama softwares have been evolved over the last decade. A comparison of commonly used panoramic softwares is provided in (Sharma et al. 2019). Image stitching algorithms have been used in many applications: aerial and satellite image mosaics (Wang et al. 2014; Truong et al. 2018; Tsai and Huang 2004), SAR image mosaicking (Benoit and Thierry 1995; Kwok et al. 1990) video frame
& Surendra Kumar Sharma [email protected] 1
Urban and Regional Studies Department, Indian Institute of Remote Sensing Dehradun, Dehradun, Uttarakhand, India
2
Department of Civil Engineering, Indian Institute of Technology Roorkee, Roorkee, Uttarakhand, India
mosaicking (Majumdar et al. 2002), 3D image stitching, virtual tourism, augmented reality (Masoud et al. 2016), and real state are few of them. Generally, image stitching algorithms consists of two main steps: image alignment and image blending. The development of image stitching technology typically depends on the innovations of these two aspects. Image alignment is used to obtain the motion relationship by detecting and matching feature points across two images or multiple images. It directly associates with the speed and success rate of im
Data Loading...