The observation of low-SNR moving targets has been a hot spot for a long time. A dynamic programming (DP) and track matching algorithm has been developed for the detection of subpixel-sized, low-SNR targets observed by mosaic imaging sensors. The primary advantages of this technique are its sensitivity to weak targets along with its robustness to target maneuvers and sensor instabilities. Enhancement is achieved by performing directional weight operation precede to DP and target data association. Experiments have been conducted and it is shown that low-SNR moving targets with maneuver and maximum speed of 1 pixel/frame can be detected with high probability.