基于Boosting学习的靶子自动检测算法研究

提出了一种实弹射击演习中的靶子自动检测方法，适合于野外复杂背景下的靶子图像处理，所有图像由固定在枪柄上的摄像头统一采集。利用经典的Boosting学习算法，将图像中靶子的特征提取出来，对图片的训练集进行训练，并从训练中学习到一个强的分类器，从而实现了靶子所在区域的自动化检测。并将方法用于一个具体实例，实验结果证明了所提方法的有效性。

Research on algorithm of automatic detection of target based on Boosting

This paper presents a new method of automatic detection on target from image captured by outdoor firing training system with complex background.All the images are picked up by a camera on pikestaff.Boosting learning algorithm is used here to extra the features of target from the images of target.And then train the training samples to gain a strong classifier,which can help to detect the location of target more accurately.We try to apply the method to a relevant example,and the result of the trial indicates the validity of the method.