复杂背景下红外小目标检测流水线快速预处理算法  

Pipeline Pretreatment for Infrared Weak-target Detection under Complex Background

作　　者：彭鼎祥[1] 汤心溢[1] 王世勇[1] 

机构地区：[1]中国科学院上海技术物理研究所,上海200083

出　　处：《半导体光电》2006年第1期75-78,共4页Semiconductor Optoelectronics

摘　　要：提出了一种实现红外小目标检测的流水线预处理快速算法:在图像增强中,利用中值滤波和Laplace滤波抑制干扰和噪声;在阈值分割中,通过子图像动态阈值分离背景和目标。同时叙述了硬件流水线实现滤波和阈值的具体步骤和组织方式,做到了图像数据和处理结果的同步输入同步输出,最后给出实验结果。A progressed algorithm is proposed,i, e. ,the pipeline pretreatment for the weaktarget detection. The disturbances and noises are eliminated by using median filter and Laplace filter in image enhancement module. The background and the target are divided in dynamic threshold module. At the same time, a effective method is proposed to achieve this algorithm on the hardware platform,in order to implement synchronous image data input and output.

关 键 词：红外小目标 Laplace滤波 中值滤波 动态自适应阈值 

分 类 号：TN216[电子电信—物理电子学]