To detect dim point targets in IR backgrounds, it is often necessary to examine targets in more than one image. A dynamic programming algorithm is a useful technique for a Track Before Detect (TBD) architecture implementation, designed to track and detect dim maneuvering targets from an image sequence under low SNR conditions. We will present a TBD architecture which combines a subpixel velocity module for detecting targest moving at subpixel velocities from frame to frame and a DPA module which integrates the signal for maneuvering targets, varying from straight line flight. The DPA does a search over all the possible state sequences, marking possible tracks by scanning each pixel in each frame and determining where it was likely to originate from in the previous image, assuming it is the true target. Each transition receives a score based on its probability of being a target track. The scores are functions of the pixels intensity, transition velocity and direction and are given while considering their surroundings and a-prior restrictions such as the allowed maneuvering. Finally, we represent a CFAR module, which allows us to determine if a target is present, and if so, where the target is present, and if so, where the target is. Data to be tested will include both real data and data which has been preregistered to minimize vibrations. The effect of vibrations will be discussed.