The automation of rotorcraft low-altitude flight presents challenging problems in control, computer vision and image understanding. A critical element in this problem is the ability to detect and locate obstacles, using on-board sensors, and modify the nominal trajectory.

There are many proposed vision based methods to perform obstacle detection and avoidance for autonomous or semi-autonomous vehicles. All methods, however, will require very high processing rates to achieve real time performance. A system capable of supporting autonomous helicopter navigation will need to extract obstacle information from imagery at rates wrying from ten frames per second to thirty or more frames per second depending on the vehicle speed. Such a system will need to sustain billions of operations per second. To reach such high processing rates using current technology, a parallel implementation of the obstacle detection/ranging method is required. This paper describes an efficient and flexible parallel implementation of a multisensor feature-based range-estimation algorithm, targeted for helicopter flight, realized on both a distributed-memory and shared-memory parallel computer. 17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION OF REPORT OF THIS PAGE

No abstract

No abstract

The automation of rotorcraft flight at low altitudes requires on-board acquisition of information about the location of objects in and around the flight path of the vehicle. Due to the covert nature of certain missions, it is desireable to use passive sensors to acquire the range information Range information about the objects can be computed by combining the optical flow resulting from the sensor motion with vehicle states from an inertial navigalion system (INS). However, the range information provided by a single passive sensor i s sensitive in the direction of flight referred as the focus of expansion This problem can be overcome by the use of more than one sensor, referred to as stereo methods.The performance of stereo and motion methods of passive ranging is compared and a recursive approach is described for processing a sequence of stereo images which will be the basis for an integrated stereo and motion method to provide more accurate range information than by using a single passive sensor. Results based on motion sequences of stereo images are presented. The approach presented will also be applicable to other autonomous systems and in robotics.