A Feature-Based Small Target Detection System
Existing small target detection systems generally use the difference image between a predicted background image and an original image. This method has two disadvantages. First, to predict the background image, the size of the structural element has to be carefully selected considering the size of small targets. Second, because of blurring, clutter such as clouds can occur around the edge of the background. To deal with these problems we propose a new feature-based detection system. The proposed method selects candidate pixels with Harris corner detector and then, again selects pixels that have a higher intensity than a threshold among the candidates. After labeling the selected candidates in order to obtain the number of pixels they have, the system decides which is a small target. In an experiment, our proposed method gave better results than the existing methods.
KeywordsHarris corner detector New White Top-Hat Labeling Histogram
Unable to display preview. Download preview PDF.
- 1.Deshpande, S.D., Er, M.H., Ronda, V., Chan, P.: Max-mean and max-median filters for detection of small-targets. In: Proc. SPIE, vol. 3809, pp. 74–83 (1999)Google Scholar
- 6.Harris, C., Stephens, M.: A combined corner and edge detector. In: Fourth Alvey Vision Conference, pp. 146–151 (1988)Google Scholar
- 8.Angel, E.: Interactive Computer Graphics. Addsion Wesley Longman (1999)Google Scholar