摘要
设计了用硬件实现的两种检测算法的原理框图．可实时对２５６×２５６的图像进行运动目标检测处理．用硬件实现了噪声的实时检测与计算，为门限值的确定提供了依据．按照所提出的算法建立的实时系统具有较强的抗噪能力和一定的自适应能力，它具有简单、实时的特点．其结构可以扩展到处理５１２×５１２的图像，适合于ＶＬＳＩ的实施．
Two types of noise insensitive motion detector with hardware implementation have been designed. The functional block diagram of two detection algorithms are discussed. In one detector, there are threshold value processing, coder, frame memory and subtracter; and in the other one, there are detection logic with a line memory and a sampled value delay. The system can perform the real time motion detection for a 256×256 image. In order to set the threshold T , real time detection and calculation of noise are implemented with hardware. The systsem has also noise immunity and adaptability. Its structure can expand in processing for a 512×512 image and is suitable for VLSI implementation.
出处
《华中理工大学学报》
CSCD
北大核心
1997年第10期80-82,共3页
Journal of Huazhong University of Science and Technology
关键词
运动检测
实时处理
信号检测
图像信号处理
硬件
motion detection
real time processing
signal detection
video monitoring
