CAJ | 학술논문

针对具有复杂背景干扰，信噪比低的红外图像目标提取问题，提出了一种结合数学形态学和自动区域生长的红外目标提取方法。首先对图像进行中值滤波和灰度变换预处理，然后利用数学形态学高帽变换去掉大面积背景干扰区域，接着采用基于局部小波系数能量特征的方法找到感兴趣区域，再通过基于窗口均值的方法自动获取种子点进行区域生长分割。实验结果表明：本文方法能有效的消除复杂背景干扰，并很好的提取出目标。
침대구유복잡배경간우，신조비저적홍외도상목표제취문제，제출료일충결합수학형태학화자동구역생장적홍외목표제취방법。수선대도상진행중치려파화회도변환예처리，연후이용수학형태학고모변환거도대면적배경간우구역，접착채용기우국부소파계수능량특정적방법조도감흥취구역，재통과기우창구균치적방법자동획취충자점진행구역생장분할。실험결과표명：본문방법능유효적소제복잡배경간우，병흔호적제취출목표。
Aiming at the infrared image target extraction with complicated background interference, a method combined with mathematical morphology and region growing is proposed. Firstly, preprocess the image by median filtering and adjusting grey level. Then remove the large background interference area using mathematical morphology transform. Next the region of interest is found by a method based on feature of local energy of wavelet coefficient. Following the seed point is obtained automatically with the method of local average and which is used in region growing segmentation. Experiment results show that this method can effectively remove the background interference and extract the target well.