| The main challenge in building wireless networks of cameras with automated detection capabilities is the high data rate. Sending all the data, even after performing standard compression, is very costly in transmission energy, and performing sophisticated vision processing locally to substantially reduce transmission rate requires high processing energy. To address these challenges, a task-driven approach, in which simple local processing is performed at each node to extract the essential information to collaboratively perform the task, has been proposed.;This dissertation presents such a task-driven approach for tracking a single object in a structured environment in the presence of static and moving occluders using a wireless camera network. To conserve communication bandwidth and energy, each camera performs simple local processing to reduce each frame to a scan line. This information is sent to a cluster head to track a point object. We assume the locations of the static occluders to be known, but only prior statistics on the positions of the moving occluders are available. A noisy perspective camera measurement model is presented, where occlusions are captured through an occlusion indicator function. An auxiliary particle filter that incorporates the occluder information is used to track the object. We investigate tradeoffs involving the tracker accuracy, moving occluder prior accuracy, number of cameras used and number of moving occluders present. We generally find that obtaining moving occluder priors may not be worthwhile, unless it can be obtained cheaply and to a reasonable accuracy.;In addition to tracking, we also consider two relevant topics, namely camera node selection and placement. Communication and computation cost can be further reduced by dynamically selecting the best subset of camera nodes. This allows efficient sensing with little performance degradation relative to using all the cameras. The minimum mean square error (MSE) of the best linear estimate of object position is used as a metric for selection. A greedy selection heuristic is proposed and it is shown to perform close to optimal. MSE metric is also used for camera placement in a simple setting. An analytical formula for this metric is presented and optimized for best camera placement. |
