数字图像中感兴趣目标的定位是一项关键的技术,它主要借助傅里叶变换对数字图像进行频域处理,然后转换到空域中进行图像的再次重构.论述了傅里叶变换的原理,通过实验对理论做了进一步的研究.通过视频监控获取多帧图像,然后对图像依次做了灰度、二值和求补的多项处理,最后提取出各帧图像中不变的特征目标.因为能量主要集中在低频分量上,只要提取显示高能量区域,就会完成对感兴趣目标的模式定位.%Interested in the digital image is a key target in the technology,it is mainly using Fourier transform in frequency domain for digital image processing,and then switch to the airspace in the reconstruction of the image again. It discusses the main principles of Fourier transform,theory through experiments done on further research. Through multi-frame images for video surveillance,and then once do a gray image, binary and complement a number of processing, the final extracts invariant features of targets from image in each frame. Because the energy is concentrated in the low frequency components,and as long as the extraction displays high-energy area,will complete the pattern of interest in target positioning.
展开▼
