为了尽可能降低噪声的影响和保留图像的边缘信息,以确保配药机器人的机械手能快速准确地抓取药瓶完成整个配药过程,结合了遗传算法和Canny边缘检测算子的优点,首先采用半像素插值的方法将3 ×3模板扩展成5×5模板,再利用图像梯度方差作为判据对图像进行四叉树分块,然后对每个子块采用二次寻优遗传算法再反复迭代求出最佳阈值,最后利用该阈值来对边缘进行提取.实验结果表明:与传统的算法相比,该算法的检测效果更好,能有效地减少假边缘点出现的概率,保证了机械手的抓取动作顺利进行.%In order to reduce the influence of noise as far as possible and maintain the image edge information to ensure the dispensing robot manipulator can quickly and accurately grasp the bottle to complete the whole process of dispensing,the advantages of genetic algorithm is combined and Canny edge detection operator.This algorithm first use half pixel interpolation method to expand template 3 × 3 to template 5 × 5,then use image gradient variance as a criterion to do image block,then for each sub-block using quadratic optimization genetic algorithm to repeat iteration to find the best threshold value,finally using threshold value to extract the edge.The experimental results show that,compared with traditional algorithm,the effect of this algorithm is much better,can effectively reduce the false edge points probability,assure the mechanical hand grasping movements smoothly.
展开▼
