【摘要】 设计一个基于时域滤波的前景检测算法,实现实时系统中前景检测和图像去噪的功能。在时域中对每个点进行前景、背景分析,从而检测出所有的前景点,并且对生成的前景、背景二值图像进行分析,过滤由噪声引起的虚景,从而较好地检测出图像中的有效前景。这种算法具有较小的时间复杂度,可以较好地应用于实际的系统中。更多还原

【Abstract】 Proposes an algorithm based on the time-domain filtering of foreground detection to achieve real-time system foreground detection and image denoising function.It is made foreground analysis and background analysis for each point in time domain to detect all the former attractions.Moreover,analyses the binary images of foreground and background to filter the virtual scene caused by the noise,which can better detect the effective foreground in the image.At the same time,the algorithm has lower time complexity and can be better applied in practical system.更多还原