Composed of imaging cameras and image processor, binocular vision system can be installed on unmanned systems such as robot, UAV and AGV, improving their capabilities in aspects of environmental sense, visual distance measurement, obstacle detection, wire detection, obstacle avoidance, direction optimization, terrain following, intelligent tracking, light stream positioning and etc.
Imitating human eyes, the system can output the distance data of any target within the field of vision according to the triangulation principle.
The binocular distance measurement module designed by QUANHOM is based on FPGA’s stereoscopic imaging technology that effectively shrinks the processor and lowers the power consumption of the device by assigning large number of simple, repetitive calculation to the hardware and making the software only assume complex logic operations.
Going though distortion and geometric corrections as well as 3D calculation, the images captured by the module contain distance data of all pixels in the field of vision which could be measured in three-dimensional coordinates. It helps the module to recognize the relative positions between an obstacle and the two cameras on a real-time basis by using wide dynamic imaging technology.
- 1.Adaption to scenes of wide dynamic range
- 2.High real-time data transaction
- 3.Three pairs of cameras in maximum could be connected to one processor at a time
- 4.Obstacle detection
- 5.Circumvention and guidance
- 6.Terrain following
- 7.Customized following
- 8.Visual odometer
Imaging resolution 640×480 Field of vision Forward/Backward
27.7° x 20.9° ±5%
52.4° x 40.5° ±5%
Distance measurement accuracy Forward/Backward
≤5%@5m ， ≤20%@10m
≤10%@5m ， ≤20%@10m
Minimum distance measured Downwards
Maximum distance measured 25m Processor size 100mm x 100 mm x 20mm power consumption <3W