PROJECT TITLE :
Supervised Segmentation of Remote Sensing Image Using Reference Descriptor
During this letter, we tend to propose the utilization of a unique feature illustration known as reference descriptor (RD) for supervised remote sensing image segmentation. Completely different from traditional low-level image features like color, shape, and texture, which are directly extracted from a picture, RD describes a data sample by its similarities to the exemplar knowledge in a very reference set and may be a higher level feature illustration of the data sample. Experiments show that comparing with segmentation using low-level image features, RD is more sturdy against intraclass variation of land cowl sort. Using RD can improve the accuracy in a very supervised segmentation (classification) framework, and superior performance is observed compared to other ways on totally different image information. Similarly, compared with the competing methods, an RD-based technique is additional economical.
Did you like this research project?
To get this research project Guidelines, Training and Code... Click Here