一种改进的基于OTSU自适应阈值的机器人数字路标图像模糊增强算法

在机器人路径规划中，机器人数字路标识别是很重要的，图像的预处理会影响识别结果。图像增强技术是提高预处理结果的一种有效方法，模糊图像增强算法是目前广泛使用的一种增强算法。针对Pal模糊图像增强算法在隶属函数的定义和渡越点选择上的缺点，提出了一种改进的模糊增强算法。本算法首先使用OTSU算子自动选择最佳阈值，解决渡越点需要人工设置的缺点，并消除选择的随机性。然后修改模糊增强算法的核心隶属函数式，解决了图像像素的低灰度值被硬性设置为0的缺陷，从而改善了图像信息损失的问题。最后，将改进的算法用于处理Pioneer Ⅲ机器人的数字路标图像。实验结果表明，与现有的模糊增强算法相比，提出的算法可以取得好的效果，且提高了运算速度，具有一定得实用性和推广性。

An Improved Fuzzy Enhancement Algorithm Based on OTSU Adaptive Threshold for Robot Digital Roadmap

In robot path planning, the robot digital roadmap recognition is very important. The image preprocessing can affect the recognition result. Image enhancement is always used in image preprocessing to provide better input for the following automated image processing, and fuzzy image enhancement is one of an effective approach that is used widely in image pre processing. Through the research of the Pal fuzzy enhancement algorithm, it is found that there are two main bugs of the al- gorithm. The first one is the definition of the membership function, and the second is the selection of the transition point. In this paper, a novel improved algorithm is proposed by analyzing the deficiencies of Pal algorithm. First, the OTSU operator is used to automatically select the best threshold value, and the disadvantage of the transition point image information loss is solved. Then, the core membership function of the fuzzy enhancement algorithm is modified. Finally, the new algorithm is applied to process digital roadmap for Pioneer Ill robot. The experimental results indicate that the proposed image enhance ment method can achieve better performance, and has higher processing speed than now-available fuzzy image enhancement algorithms. And it also possesses the property of high practicability and generality.