Automatic Road Extraction from High-Resolution Remote Sensing Image Based on Bat Model and Mutual Information Matching

Abstract

Considering automatic approaches of road extraction have two difficulties: the first one is how to identify initial tracking position and direction automatically, and the second one is how to complete the tracking process correctly with the disturbing influence, we proposed a new method of automatic road extraction from high-resolution remote sensing image based on bat model and mutual information matching. Firstly, for determining the initial tracking position and tracking direction automatically, we proposed a fully new model, which was called bat model. It included the following aspects: region seed detection based on improved Harris detector and road seed extraction based on fully new bat algorithm. By taking into account the context features of candidate road seeds, the bat algorithm simulated bat behavior to search all forward paths from current position and determine the best moving direction. Secondly, for road tracking, we designed a new method to get initial reference template automatically and proposed road tracking based on mutual information matching. According to the experiment results, the proposed method can determine initial tracking position correctly, and complete the tracking process automatically even if the road represents a shape of ribbon with a big bending and the disturbing influence

Lane-Level Road Extraction from High-Resolution Optical Satellite Images

“…Edge information is usually expressed in two forms: edge points (Li et al [42]) and edge lines. Edge points are a discrete form of information sensitive to local mutation points, and the regularity of edge points is difficult to characterize.…”

“…This paper considers the road interior (road-marking lines, edges of mobile vehicles, central median dividers, and shaded edges of flat-roofed buildings), road edges (lined by trees and edges of parked vehicles), and objects near the road (edges of buildings) to usually have the same or a similar direction to that of the road and predicts the tracking direction by capturing the semantic relationships between shadows, trees, motor vehicles, road indicators, and other features. Edge information is usually expressed in two forms: edge points (Li et al [42]) and edge lines. Edge points are a discrete form of information sensitive to local mutation points, and the regularity of edge points is difficult to characterize.…”

Real-time Intelligent Alert System on Driver�s Hypo-Vigilance Detection Using Template Matching Technique

“…In spite of the accuracy of medical methodology, it's complicated and need certain environment which made it hard to generalize. The research "An Evaluation on Various Vision-based Fatigue Driving Detection Methods" which was conducted compared four methods and nine parameters fatigue driving results, and PRECLOS especially P80 detection method shows it's superiority [27] [28].…”

Digital Image Stabilization Based on Harmony Filter

“…Harris detector which has been successfully applied to different fields recently [16,17] is proposed by Harris and Stephens [18] based on the auto-correlation matrix. .…”