Automatic oil tank detection algorithm based on remote sensing image fusion

Automatic target detection like oil tank detection is one of the important domains in image processing. Which could be used for disaster monitoring, oil leakage, etc. Because of differences of objects spectrums, image fusion may be used to improve the features while remain image resolution In this paper automatic oil tank detection algorithm improves recognition ration, and computational time is reduced, using improved Romanized Hough Transform, improved Canny algorithm and Fast Template Matching algorithm. Panchromatic image and multi-spectrum images of SPOT-5 satellite are used as the example. The experiment shows the algorithm reduces computational time of RHT and Fast Template Matching, and the ratio of oil tank detection exceeds 85%.