Intensity invariant nonlinear correlation filtering for detection of camouflage patterns

• 主要作者: Heng, Boon Hua.
• 其他作者: Qian Kemao
• 格式: Final Year Project
• 語言: English
• 出版: 2012
• 主題: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Pattern recognition; DRNTU::Engineering::Computer science and engineering::Computing methodologies::Simulation and modeling
• 在線閱讀: http://hdl.handle.net/10356/48552
• 機構: Nanyang Technological University

Trying it to spot a camouflaged target in a forest far away is hard enough for us. Our naked eyes can only process these much to search and location an unknown target. Hence we need an equipment/ways to provide reliable and accuracy solution to handle this task. Accuracy for most pattern recognition techniques is a problem when searching for target in a camouflaged environment under illumination conditions. Different illuminations conditions may cause the correlation values to vary. Consistent is also important so that even with different combination of illumination conditions and camouflaged pattern, results remain similar and not random. This technique does need to know the exact camouflaged pattern but the class of the pattern. In this report, we will look into an algorithm method based on orthonormal vector space basis representation to detect camouflaged in natural environments by Henri H. Arsenault and Pascuala Garcia-Martinez. We will analysis and experiment to determine how efficient it is by simulating the method.