基于PCNN的桨叶图像提取及锥度测量

摘要

针对目前对直升机桨叶共锥度测量难度大、精度低、无法对桨叶逐点测量的缺点,提出了一种应用视觉系统测量旋翼共锥度的新方法,给出了系统的总体安装方案并对系统的成像误差精度进行了分析,由于旋翼桨叶处于高速旋转状态需要实时处理的特点,提出了一种基于脉冲藕合神经网络(PCNN)的分割方法,在设计好的旋翼桨叶模型上进行了相关实验,通过与其它分割算法的对比,验证了算法的具有很高的识别率;实验结果表明:基于PCNN的分割方法具有较高的精度,适用于实时的旋翼锥度测量系统.%According to the problems for helicopter blades pyramid angle measurement, such as great difficulty of the measuring, low accuracy and thoroughly examination hardly, this paper presents a new method applied to helicopter blade pyramid angle measurement by using vision system, the overall design scheme of the system is offered and the imaging error is analyzed.Because the rotor blades rotated at a high speed and need real-time processing, proposes a image segmentation method based on Pulse Couple Neural Network (PCNN), the related experiment to the rotor blades model validates the PCNN algorithm has the advantage in recognition rate compared with other segmentation algorithms.The experimental results show the method has higher accuracy and suitable for blades pyramid measurement system.