How to accurately extract the characteristics of specific areas of combat vehicles operations from a complex background image was studied in this paper. Having illustrated the limitations of the traditional ways in Identifying combat vehicles in the complex background image ? A new method based on morphological opening operation with multi-threshold segmentation algorithm was put forward. Firstly, analysis the image intensity and select the appropriate threshold value as well as do the morphology opening operation. Then split the specific area according to local features combat vehicles is significant again in order to segmentation and extract the feature of Combat Vehicles from the image of complex background, finally use the tracking box to mark the region of interest. The result of experiment shows that the new method works well in identifying combat vehicle in the complex background image.%对如何从复杂的背景图像中准确提取作战车辆特定区域的特征进行研究,阐述传统识别方法应用于复杂背景图像中作战车辆特征提取的局限性,提出基于作战车辆形态学开运算结合多阈值处理的分割算法.首先,通过图像强度的分析选取适当阈值对其进行形态学运算,然后根据作战车辆局部特征明显,对特定区域进行再分割,实现对复杂背景图像中作战车辆特征的准确分割与提取,最后使用追踪框将感兴趣区域做标记.实验结果表明本方法对复杂背景的作战车辆识别是有效的.
展开▼
