23 January 2018 Adaptive target binarization method based on a dual-camera system
Author Affiliations +
Abstract
An adaptive target binarization method based on a dual-camera system that contains two dynamic vision sensors was proposed. First, a preprocessing procedure of denoising is introduced to remove the noise events generated by the sensors. Then, the complete edge of the target is retrieved and represented by events based on an event mosaicking method. Third, the region of the target is confirmed by an event-to-event method. Finally, a postprocessing procedure of image open and close operations of morphology methods is adopted to remove the artifacts caused by event-to-event mismatching. The proposed binarization method has been extensively tested on numerous degraded images with nonuniform illumination, low contrast, noise, or light spots and successfully compared with other well-known binarization methods. The experimental results, which are based on visual and misclassification error criteria, show that the proposed method performs well and has better robustness on the binarization of degraded images.
© 2018 Society of Photo-Optical Instrumentation Engineers (SPIE)
Jing Lei, Ping Zhang, Jiangtao Xu, Zhiyuan Gao, Jing Gao, "Adaptive target binarization method based on a dual-camera system," Optical Engineering 57(1), 013105 (23 January 2018). https://doi.org/10.1117/1.OE.57.1.013105 . Submission: Received: 30 August 2017; Accepted: 22 December 2017
Received: 30 August 2017; Accepted: 22 December 2017; Published: 23 January 2018
JOURNAL ARTICLE
15 PAGES


SHARE
Back to Top