Cell Coverage Estimation Based on Structured Random Forests Edge Detection and Boundary Tracking
Download as PDF
DOI: 10.23977/AICT2020003
Corresponding Author
ABSTRACT
To estimate the cell coverage accurately and effectively, an approach based on the structured random forests with mutual information and boundary tracking is proposed. And with the structured random forest, which is improved with mutual information to enhance the mapping function from the structured information to discrete labels, the salient and sematic edges are detected. On the basis of detected edges, the cell contours are found using boundary tracking algorithm. And the experiments imply the effective of the proposed method.
KEYWORDS
Edge Detection; Boundary Tracking Algorithm; Structured Random Forests