Abstract:

    Adaptive tuning methods for digital camera parameters are proposed to increase the robustness of machine vision system under various illuminations. Camera parameters are online estimated first; then, analog to digital conversion level, gain, black level and exposure time are tuned according to the environmental illumination to guarantee that the average intensity of binary image is within a predetermined range. Experiments demonstrate that the method of adaptive tuning is helpful to capture canonical images in a large scale of illumination variation. The proposed method can weaken most influences of various illuminations on the machine vision system. The precision of subsequent position measurement is improved and automatic running of the machine vision system is guaranteed under various illuminations.

Key words: automatic control technology, machine vision, illumination, digital camera, visual servo