Four updated video guidance sensor (VGS) systems have been proposed. As described in a previous NASA Tech Briefs article, a VGS system is an optoelectronic system that provides guidance for automated docking of two vehicles. The VGS provides relative position and attitude (6-DOF) information between the VGS and its target. In the original intended application, the two vehicles would be spacecraft, but the basic principles of design and operation of the system are applicable to aircraft, robots, objects maneuvered by cranes, or other objects that may be required to be aligned and brought together automatically or under remote control.
In the first two of the four VGS systems as now proposed, the tracked vehicle would include active targets that would light up on command from the tracking vehicle, and a video camera on the tracking vehicle would be synchronized with, and would acquire images of, the active targets. The video camera would also acquire background images during the periods between target illuminations. The images would be digitized and the background images would be subtracted from the illuminated- target images. Then the position and orientation of the tracked vehicle relative to the tracking vehicle would be computed from the known geometric relationships among the positions of the targets in the image, the positions of the targets relative to each other and to the rest of the tracked vehicle, and the position and orientation of the video camera relative to the rest of the tracking vehicle.
The major difference between the first two proposed systems and prior active- target VGS systems lies in the techniques for synchronizing the flashing of the active targets with the digitization and processing of image data. In the prior active-target VGS systems, synchronization was effected, variously, by use of either a wire connection or the Global Positioning System (GPS). In three of the proposed VGS systems, the synchronizing signal ould be generated on, and transmitted from, the tracking vehicle.
In the first proposed VGS system, the tracking vehicle would transmit a pulse of light. Upon reception of the pulse, circuitry on the tracked vehicle would activate the target lights. During the pulse, the target image acquired by the camera would be digitized. When the pulse was turned off, the target lights would be turned off and the background video image would be digitized. The second proposed system would function similarly to the first proposed system, except that the transmitted synchronizing signal would be a radio pulse instead of a light pulse. In this system, the signal receptor would be a rectifying antenna. If the signal contained sufficient power, the output of the rectifying antenna could be used to activate the target lights, making it unnecessary to include a battery or other power supply for the targets on the tracked vehicle.
The third proposed VGS system could include either passive or active targets. This system would include two or more video cameras and associated digitizing and digital image-processing circuitry on the tracking vehicle for acquiring stereoscopic pairs of images of the targets on the tracked vehicle. At distances beyond the normal VGS operating range (that is, at distances so great that the target images would merge into a single spot of light on each camera focal plane), a VGS system operating in its normal short-range mode could determine the direction to the tracked vehicle but could not determine the distance to, or the orientation of, the tracked vehicle. However, in such a situation, this proposed system would determine the distance to the tracked vehicle by use of the known geometric relationships of stereoscopy — provided, of course, that the distance were not so great as to bring the stereoscopic disparity below the minimum useful level.
The fourth proposed system would be an active-target VGS system in which synchronization would not involve the transmission of pulses from the tracking vehicle. Instead, the target lights would be flashed at a repetition rate of 5 Hz governed by a free-running oscillator on the tracked vehicle. Each flash period would include a lights-on interval of 3/60 of a second (corresponding to three video fields at standard video frame rate of 30 Hz at two fields per frame) and a lights-off interval of 9/60 of second. The system would digitize two pictures in a row, subtract them, and look for the expected target pattern in each synthetic image generated by the subtraction. If the target pattern were thus found, then the flash timing would be known to within one field. If the target pattern were not found, then the time of each picture would be advanced one frame (two fields – 1/30 of a second) relative to the beginning of a 5-Hz processing cycle and the aforementioned actions repeated. This process would quickly bring the digitizing and
data-processing circuitry into synchronism with the flashing of the targets.