论文部分内容阅读
飞机目标识别方法研究在军事上具有重要意义。针对飞机目标图像数据的复杂性,提出了基于监督局部保持投影(SLPP)算法的一种目标识别方法。首先,利用PCA对飞机图像数据进行预降维,由此克服小样本问题;其次利用SLPP算法对飞机图像进行维数约简;最后,采用最近邻分类器进行飞机类型分类。通过在真实飞机图像数据库上的实验结果表明,该方法有效地提高了飞机识别的正确率。利用该方法能够有效地进行飞机目标识别。
The research of aircraft target recognition method is of great military importance. Aimed at the complexity of aircraft target image data, a target recognition method based on Supervised Local Retained Projection (SLPP) algorithm is proposed. Firstly, the PCA is used to pre-dimension the aircraft image data to overcome the small sample problem. Second, the SLPP algorithm is used to reduce the dimension of the aircraft image. Finally, the nearest neighbor classifier is used to classify the aircraft type. The experimental results on the real aircraft image database show that this method effectively improves the accuracy of aircraft recognition. This method can effectively identify aircraft targets.