基于区域分割的噪声抑制及红外目标检测

NOISE SUPPRESSION AND DETECTION OF INFRARED TARGET BASED ON REGION SEGMENTATION

提出一种新的基于区域分割与邻域梯度差平方累积最小(MCSE)原则的自适应滤波算法.该算法对各种加性、乘性噪声的滤波性能优异且均衡,适合滤除图像序列中的复杂噪声.滤波后,采用改进光流方程算法(OFE)估计图像序列的运动场,为避免传统算法在图像补偿时引入的误差,本文直接对运动场进行全域运动补偿,然后在运动场中引入一种新的集成检测算法对微小目标的运动矢量进行检测.理论分析及试验表明,在复杂噪声干扰下,本算法依然能够在具有全域运动的图像序列中检测微小目标.

A new adapt filtering algorithm was proposed based on region segmentation and minimum cumulation-squared error （MCSE） of gradients in neighborhood. This algorithm is good and balanced for various plus noise, multiplicative noise, or complex noise. After filtering, the improved optical flow equation （OFE） was adopted to estimate the motion field. To avoid the compensation error of traditional algorithms, global motion in images motion field was compensated. A new integrated detection algorithm was proposed on motion field to detect the small object. Theoretical analysis and experiments demonstrate that this algorithm can detect dim small objects in image sequence with global motion under strong complex noise.