智能机器人视觉传感集成体系

Inspired by the architecture of the general intelligent robots，which is always structured in levels and module, a new point of view that machine vision should also be structured in levels and module has been put forward aiming at the argument and the problems existing in the different views for machine vision research. This vision structure system can to be well satisfied with the requirements of the intelligent robot whole system to have a compromise between the general and special tasks as a vision sub-unit. .After discussing and analyzising the level architecture and various module functions in detail. A machine vision integrated sensing system, which included a laser imaging radar, 3D surface sensor and two revolving CCD cameras, had been built and implemented perfectly, in which a synchronized spot scanning triangulation method will act as 3D imaging sensing within the near range with precision data, a laser scanning range finder will play the role at the far field to get the coarse imaging for navigation and object movement detection, a CCD camera will play the role to get the intensity image for object recognition such as color, gray and texture information on the object surface. They will be united in space and time, and all the information can be matched and complemented each other. By preliminary processing, several feature databases will be established, involving 3D object surface shape database at near field、range image database at far field、movement database from a long period、intensity image database、range and speed database for feature points et. al., and they will be optimized and communicated each other. Further processing can be provided if necessary, and the system speed、 space resolution、range accuracy and the field of view of this sensing structure can be actively modified and compromised

从机器视觉面向应用、面向行为出发，开展视觉传感集成体系研究，以快速、精确、全面地感知场景信息为目的，克服遮挡和阴影效应、适应远场及近场特点；研究这一传感体系中不同信息的协调、匹配、标定、冗余量剔除、误差修正及恢复等方法。该项研究将为推动机器视觉研究的发展，提高机器的智能化水平奠定基础。