【摘要】 将提升小波的提升项作为可调参数,通过训练使其具有目标的特征,利用提升后的滤波器提取出目标,从而实现目标的检测。仿真实验表明:此方法比传统的模板匹配方法速度大大提高。为使基于小波变换的目标检测算法能应用在实际检测跟踪系统上,分析在DSP系统中实现基于小波变换的目标检测算法的可能性和优势,并给出以双DSP为核心器件TMS320C6203的实时图像处理系统,从而为目标检测算法的硬件实现奠定实验基础。更多还原

【Abstract】 The lift item of lifting wavelet can be seen as an adjustable argument.By training,let it have the character of target.The target is got from the filter constructed after lifting,thus the detecting is realized.The emulated experiment results show that this algorithm provides a speed realization compared with the traditional template matching method.In order to apply the target detecting algorithm based on wavelet in the detecting and tracking system,the authors explore the possibility and advantages of implementing this algorithm based on DSP system,and give the dual-core DSP devices for real-time TMS320C6203 image processing system,thus the experimental basis for the hardware implementation of target detecting algorithm is established.更多还原