基于图像识别的哑弹体特征提取方法研究

The Approach of Detection Dumb Bomb Feature According to Image Identify

针对哑弹体的图像轮廓,分析了关于直线和边缘点的几种提取方法,找出了一种适合提取弹体边缘的提取方法——分裂算法。此算法的优点是可以比较精确地找到曲线的转折点,并用这些转折点来描述曲线,对于直线和折线组成的图形有较好的逼近效果。

For the image outline of dumb bomb, several methods of the detection about line and edge point are analyzed, found a method which fit detection of bomb outline - splitted calculate way. It cut off a lengh curve into some small curves, and use lines or broken lines proximate these curves until they are satisfy standard. The advantage of this method is that it can find out transition point of curve accurately, And description curve by these transition. There is a good approach effect for graphical which is consist of lines and broken lines especially.