Point target tracking in a whitened IR sequence of images using dynamic programming approach

Abstract

In this paper, a novel tracking system based on the Track Before Detect Approach (TBD) is introduced. Each IR sequence is preprocessed first by using a whitening algorithm. This stage is used to reject clutter and emphasize targets. Afterwards, a Dynamic Programming Algorithm (DPA) is used for numerous frames in the sequence. The algorithm, a derivative of the well known Viterbi Algorithm, gives each pixel in the image a score based on the current frame and the previous one. By doing so, the temporal behavior difference between targets, clutter and noise is utilized to distinguish between them; we give scores accordingly. At the end of this stage, after the last frame of the IR sequence has been processed, the pixel with the highest accumulated score is chosen as the Target, and its path is found. The paper deals with the different issues characterizing the system, enabling it to have versatility over a wide range of scenes. Future work will involve the use of the system for tracking of targets in hyperspectral cubes.

Spatial and temporal point tracking in real hyperspectral images

“…Parts of this study have been published previously by our group: we will, therefore, refer extensively to those publications. Algorithms for target detection in single hyperspectral cubes are described in Raviv and Rotman [20], the details of the VF and of the generation of the hyperspectral movie are presented in Varsano et al [13], and the DPA is described in Nichtern and Rotman [14]. In this article, we present an overall integration of the system; in particular, the article analyzes the integration of the VF and the DPA and provides an overall evaluation of the system.…”

“…Type 1 signals are the most problematic, since the DC of such signals does not have an overall fit with a linear model, but depends on piecewise matching of the DC to windows sizes, as explained below. Figures 11,12,13,14,15,16,17,18,19, and 20 illustrate the algorithm's operation on the various signal types, with and without an implemented target. In each case, the DC signal and the estimated variance values (calculated after subtracting the estimated DC from the signal) are also plotted.…”

“…Our system uses algorithms from two different areas, target detection in hyperspectral imagery [1][2][3][4][5][6][7][8][9] and target tracking in IR sequences [10][11][12][13][14][15][16][17][18][19]. Numerous works have addressed each of these problems separately, but to the best of our knowledge, to date no attempts have been made to combine the two fields.…”

“…The filter detects the presence of targets from the temporal profile of each pixel, while suppressing clutter-specific influences. Finally, a track-before-detect (TBD) approach is implemented by a dynamic programming algorithm (DPA), to perform target detection in the time domain [14][15][16][17]19]. Performance metrics are defined for each step and are used in the analysis and optimization.…”

A Survey of Motion-Based Multitarget Tracking Methods

“…The whole performance analysis of DP techniques was presented in [174,175]. In IR image tracking [176,177], DP algorithm gives each pixel in the image a score based on the current frame and the previous one. By doing so, the temporal behavior difference between targets, clutter and noise is utilized to distinguish between them; the algorithm gives scores accordingly.…”