Improved target detection algorithm using Fukunaga-Koontz transform and distance classifier correlation filter

Abstract

Often sensor ego-motion or fast target movement causes the target to temporarily go out of the field-of-view leading to reappearing target detection problem in target tracking applications. Since the target goes out of the current frame and reenters at a later frame, the reentering location and variations in rotation, scale, and other 3D orientations of the target are not known thus complicating the detection algorithm has been developed using Fukunaga-Koontz Transform (FKT) and distance classifier correlation filter (DCCF). The detection algorithm uses target and background information, extracted from training samples, to detect possible candidate target images. The detected candidate target images are then introduced into the second algorithm, DCCF, called clutter rejection module, to determine the target coordinates are detected and tracking algorithm is initiated. The performance of the proposed FKT-DCCF based target detection algorithm has been tested using real-world forward looking infrared (FLIR) video sequences.

Improved training for target detection using Fukunaga-Koontz transform and distance classifier correlation filter

“…In target detection and tracking applications, the target may temporarily go out of the field-of-view in some sequences leading to the reappearing target detection problem [1,2]. The target(s) of interest may reappear at a later frame.…”

“…K-means algorithm is employed to cluster the training frames and the number of training chips is depending on the number of clusters such that training chip for each cluster. Then, the algorithm employed FKT and DCCF for detection and tracking similarly to [2].…”

“…The DCCF is used in the classification stage to improve the detection and discrimination capability [3]. The FKT and DCCF filters are trained with previous signatures of target images with expected size and orientation variations of the target [2]. The training part of the DCCF continues before the target disappears from the field-of-view by using the K-means algorithm.…”

“…The target and candidate classes which are obtained in this step are employed later in the testing part. When the target goes out of the field-of-view, the next consecutive input scenes are correlated with FKT and tentative target regions are determined [2]. Then, these candidate targets are used to identify target types and reject clutters based on a threshold.…”

“…In [2], the training part is accomplished manually, which means that a user select the training chips from the training sequences. In addition, the author selected the number of training chips randomly.…”