摘要
利用Mallat塔式分解的优点,提出一种基于Mallat分解的序贯相似性检测算法(SSDA),这种算法在保证匹配准确性的基础上,大大缩短了匹配时间。实验证实,本匹配方法具有一定抗几何失真的能力,优于常规的匹配方法,能够快速准确地检测出织口的位移变化,解决了应用图像处理方法消除织机开车稀密路系统中匹配时间过长的难题。
An automatic approach of image matching method based on Mallat pyramids decomposition is developed. This approach can not only garantee precision of image matching, but also sharply reduce matching time. As proved, this matching method is still effective despite geometrical distortion of the fabric, It is superior to general image matching method and will shorten matching time when eliminating set marks by image processing method.
出处
《东华大学学报（自然科学版）》
CAS
CSCD
北大核心
2006年第6期86-88,共3页
Journal of Donghua University(Natural Science)
