Evaluating Performance of Feature Extraction Methods for Practical 3D Imaging Systems
Smart cameras are extensively used for multi-view capture and 3D rendering applications. In the authors' study, they investigate the behaviour of SIFT, SURF and ORB through simulations of practical issues and evaluate their performance targeting 3D reconstruction (based on epipolar geometry of a stereo pair). Their experiments are carried out on two real-world stereo image datasets of various resolutions. Their experimental results show significant performance differences between feature extractors' performance in terms of accuracy, execution time and robustness to blur, lens distortion and thermal noise of various levels. Eventually, their study identifies suitable operating range that helps other researchers and developers of practical imaging solutions.