Object Localization Using Linear Adaptive Filters

We present a novel approach to localization of objects in clutter images with the use of linear adaptive filters in a two-object classifier: target object versus clutter object. An automatic optimized feature extraction processing is suggested to generate two pair of models: “target” and “clutter” models from training image databases, and “clutter-like-target” and “target-like-clutter” models from positive and negative detection errors examples respectively. Experimental results obtained on testing database of known acquisition system containing “face” and “non-face” objects show that the proposed approach outperforms other literature reported results both in term of detection rate and false alarm rate.