Unmanned Aerial Vehicle Remote Sensing Image Segmentation Method by Combining Superpixels with multi-features Distance Measure

6th Annual 2018 International Conference on Geo-Spatial Knowledge and Intelligence

Huang, Liang^1^2 ; Song, Jing^1 ; Yu, Xueqin^3 ; Fang, Liuyang^4

Faculty of Land Resource Engineering, Kunming University of Science and Technology, Kunming

650093, China^1

Surv. and Mapping Geo-Info. Technol. Research Center on Plateau Mountains of Yunnan Higher Education, Kunming

650093, China^2

Kunming Surveying and Mapping Institute, Kunming

650051, China^3

Broadvision Engineering Consultants, Kunming

650041, China^4

关键词: Change detection;    Distance measure;    Iterative clustering;    Multi features;    Remote sensing images;    Segmentation methods;    UAV remote sensing;    Visual evaluation;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/234/1/012022/pdf
DOI  :  10.1088/1755-1315/234/1/012022

来源: IOP

【 摘 要 】

Image segmentation is the foundation and key step of object-level classification and change detection. In this paper, a segmentation method of UAV remote sensing image based on multi-features distance measure and superpixels is proposed. First, the simple linear iterative clustering (SLIC) algorithm is used to segment the unmanned aerial vehicle (UAV) remote sensing image to obtain the initial superpixels. Then the distance measures of the spectral, texture, shape and area features are used as the criterion for initial superpixels merging. Finally, merger termination when the number of regions reaches the set number. Two groups of UAV remote sensing images are selected to evaluate the experimental results through visual evaluation. The experimental results show that the proposed method can be used to aggregate objects of different scales, and the segmentation effect is satisfactory.