Frame-rate omnidirectional surveillance and tracking of camouflaged and occluded targets

Abstract

BackgroundThere has been considerable work on tracking systems, for example, see [11] [9]. Our system draws ideas from these and other earlier work. While many of the basic ideas are similar, the details are often quite different, and are what account for the systems unique abilities.Some of the major differences stem from our area of application. Our goal is to track targets in a perimeter security type setting, i.e. outdoor operation in area of moderate to high cover. We seek real-time algorithms suitable for COTS (Common-Off-The-Sheff) type of computing, and use x86 based processors. This domain of application significantly restricts the techniques that can be applied. Some of the conThis work supported in part by DARPA Image Understanding's VSAM program. straints, and their implications for our systems include:The lighting is naturally varying. We must handle sunlight filtered through trees and intermittent cloud cover. (We are not considering IR cameras, yet).Targets use camouflage, thus it is unlikely that color will add much information. Figure 3 shows an example scene with a sniper in the grass.Targets will be moving in areas with large amounts of occlusion; finding/classifying outlines will be difficult.Trees/brush/clouds all move. The system must have algorithms to help distinguish these "insignificant" motions from target motions.Many targets will move slowly (less than [1/ 60] pixel per frame); some will move even more slowly. Some will try very hard to blend into the motion of the trees/brush. Therefore frame-to-frame differencing is of limited value. Temporal adaption schemes must not add slow targets to the background.Targets will not, in general, be "upright" or isolated. Thus we have not added "labeling" of targets based on simple shape/scale/orientation models.Targets need to be detected quickly and when they are still very small and distant, e.g. about 10-20 pixels on target.Correlation, template matching, and related techniques cannot be effectively used because of large amounts of occlusion and because in a paraimage, image translation is a very poor model; objects translating in the world undergo rotation and non-linear scaling.Note that, except for the last, these are all generic problem constraints and are not dependent on the geometry of the paraimage. If a system can track under these constraints it can be used in many situations, not just omni-directional tracking in outdoor settings.We also note that, the detection phase is crucial; if targets are not detected they will not be tracked. Detection is also an area where the domain constraints make this more difficult than the situtations considered in most past papers. As a result, much of this paper (and the systems effort) is concentrated on the detection phase. Because of the camouflage and 1

Into the woods: visual surveillance of noncooperative and camouflaged targets in complex outdoor settings

“…While it may seem counter intuitive, the spatial resolution of the paraimages is greatest along the horizon, just where objects are most distant. In [28] we show that along the horizon, the resolution of an omnicamera is 4.2 pixels per horizontal degree, which is about the same as three traditional cameras with 150 degree FOV that would be needed to watch the same region. With either an omni-directional camera or many traditional cameras, objects to be tracked in a wide field of view will cover only a small number of pixels.…”

“…Frame-to-frame motion is small -a good sniper may crawl at under a tenth of a meter per minute and be motionless for minutes at a time. A video of this sniper being tracked by the Lehigh Omni-directional Tracking System (LOTS) [1,2] can be found at http://www.eecs.lehigh.edu/˜tboult/TRACK.…”

“…[2,15,29,13], so these systems support multiple background models per pixel. One such model, used in [15,29], is to fit a MOG to the given input samples.…”

“…The parametric form of the MOG distributions then can be used to classify pixels. In [2], a simpler form is used that tracks only the central values of the two primary distributions for a pixel. These papers draw mostly on intuition and insight, and do not present experiments justifying their multiple background model assumption or parameter settings.…”

Designing a Posture Analysis System with Hardware Implementation

“…This single-person tracking assumption is also made by a number of existing tracking systems [14,17,18]. Systems to track multiple people exist, both for isolated people [3,13] or people in groups [10].…”