摘要
小目标检测问题是电视跟踪领域的一项关键技术,针对海空背景下少像素的舰船目标,根据其背景不均匀以及海面噪声干扰大,舰船像素少,但是存在相对连续的目标小块区域的特征,用一定尺度的小波变换对不同子带自适应选取阈值来去除噪声干扰、抑制背景,同时结合灰度形态学滤波能很容易地检测出小目标.实验结果表明:该算法能够有效地检测出强噪声和背景干扰情况下的舰船目标.
Detection of small target with low contrast is one of the key techniques in TV tracking field. An adaptive, multi-threshold method is proposed for image denoising and background suppressing, based on different subbands. On the characters of small targets in the background of sky, sea, and sea-sky, has fewer pixels but exits a small continuous region. The gray morphology filter is employed to execute filtering to detect the true small target. The result of experiment shows that this method is able to get satisfied results in the special background.
出处
《电子器件》
CAS
2007年第3期992-994,998,共4页
Chinese Journal of Electron Devices
关键词
目标检测
背景抑制
去噪
小波变换
灰度形态学
target detection
background suppressing
denoising
wavelet transform
gray morphology
