作　　者： ; ;

机构地区： 南昌大学

出　　处： 《南昌大学学报（理科版）》 2004年第3期297-300,共4页

摘　　要： 运用启发式搜索策略提出了一种基于彩色离焦模糊图像清晰度评价的启发式调焦算法。算法采用了以2为模的Sobel算子梯度建立图像清晰度评价函数的数学模型。该评价函数具有无偏性好,单峰性强,灵敏度高,在焦平面变化趋势明显等特点。我们给出了结合启发式搜索策略的具体算法。实验结果证明,所提出的启发式调焦算法精度高,可重复性强,并对图像噪声具有良好的鲁棒性。 Focusing is one of the most important initialization work before using vidicon systems. Manual Focusing usually asks for operators to master a certain specialized knowledge, and the operations themselves are trivial. In this paper, we proposed a heuristic focusing algorithm with heuristic search strategy. In our algorithm, we build up the mathematic model of image definition evaluation with the gradient of Sobel operator modulo 2. This evaluation function has features of small deviation, single peak, and high sensitivity, apparent variation trend near focus plane. And we give the practical heuristic focusing algorithm in the paper. As experiment results show, our heuristic focusing algorithm is accurate, repeatable and robust to image noises.

关 键 词： 算法 模糊图像 启发式搜索 评价函数 彩色 鲁棒性 算子 离焦 清晰度 图像噪声

领　　域： [自然科学总论] [自动化与计算机技术] [自动化与计算机技术]