Abstract:
Automatic detection of certain landmarks, objects or other targets from multispectral aerial / satellite images is an emerging area of research with articulate demands in military and civil sector. Bridge detection from aerial images is one such key landmark that has vital importance in surveillance, disaster management and relief missions. Specifically, during floods / earthquakes, relief operations are carried out by UAVs. They can provide the damage assessment of bridges for community relief, rescue and help identifying alternative transport networks.
In this research, a multi-stage technique in light of fractal hypothesis and knowledge mining is proposed. The novelty of the project is to detect bridges from low and high oblique aerial images instead of straight down (orthogonal views). The camera and environmental parameters are assumed to be known a priori. During the first stage, corner and edge detection is carried out by using gradient covariance. Edges with more texture environs are suppressed by introducing a measure of isotropic environs and allocating large weights to strong edges and clear boundaries. For the second stage, an improved Hough transform voting scheme is used on edge image along with its energy to extract all possible lines as per the relevant description of the target orientation and length.
The target is identified by evaluating crosswise feature examination that seeks self-resemblance across the target. An efficient computation method for calculating the isotropic surround suppression is used that accelerates the proposed algorithm. Simulation results on real-world and synthetic images are used to demonstrate the effectiveness of the proposed method in automatic target detection.
