碳纤维钢针置换装置的自动对焦策略研究

Study on Auto-focus Strategy for Carbon Fiber Needle Replacement Device

为使碳纤维钢针置换装置的视觉系统实现自动对焦,提出了一种基于图像处理的自动对焦策略。该策略采用Tenengrad梯度算子和8方向的Sobel梯度算子,分别作为粗对焦和精对焦阶段的图像清晰度评价函数;提出一种基于图像降采样的改进爬山算法,在不同对焦阶段对图像采用不同次数的降采样,提高了对焦函数的灵敏度。同时,在机械臂上进行视觉系统对焦实验。结果表明:该自动对焦策略切实有效,对焦灵敏且鲁棒性强。

In order to realize autofocus in the vision system of carbon fiber steel needle replacement device,an autofocus strategy based on image processing is proposed.The strategy uses the Tenengrad gradient operator and the 8-direction Sobel gradient operator as the image sharpness evaluation function of coarse focus and fine focus respectively.An improved hill climbing algorithm based on image downsampling is proposed,which adopts different times of downsampling in different focusing stages to improve the sensitivity of focusing function.At the same time,the visual system focusing experiment is carried out on the robot arm.The results show that the autofocus strategy is effective,sensitive and robust.