Method of Target Recognition in High-resolution Remote Sensing Image Based on Visual Saliency Mechanism and ROI Region Extraction

Bing Liu,
Tingwei Chen,
Ping Fu,
Yumei Zhen,
Jeng-Shyang Pan,

Abstract


Different from those in low-resolution tasks, targets in high-resolution remote sensing recognition tasks are closed. Usually targets with higher resolution, such as oil tanks and ships bear relatively simple features and gather densely in small areas. The recognition of such targets is more dependent on the fundamental features. Ships, tanks and aircraft targets with high resolution are closed targets, which contain obvious contour and have features that are quite different from the surrounding environment. Therefore, aiming at the targets in remote sensing images with high resolution of sub-meter that is below 1m, we proposed a recognition method based on region of interest (ROI) extraction using visual saliency mechanism. Combining the result of saliency detection, MeanShift algorithm is used to image segmentation and feature extraction. Then, machine learning algorithms are applied to perform the target recognition. We achieved a recognition accuracy of more than 85% on the dataset of aircraft, oil tank and ship targets in our work.


Citation Format:
Bing Liu, Tingwei Chen, Ping Fu, Yumei Zhen, Jeng-Shyang Pan, "Method of Target Recognition in High-resolution Remote Sensing Image Based on Visual Saliency Mechanism and ROI Region Extraction," Journal of Internet Technology, vol. 20, no. 5 , pp. 1333-1341, Sep. 2019.

Full Text:

PDF

Refbacks

  • There are currently no refbacks.





Published by Executive Committee, Taiwan Academic Network, Ministry of Education, Taipei, Taiwan, R.O.C
JIT Editorial Office, Library and Information Center, National Dong Hwa University
No. 1, Sec. 2, Da Hsueh Rd. Shoufeng, Hualien 97401, Taiwan, R.O.C.
Tel: +886-3-931-7017  E-mail: jit.editorial@gmail.com