摘要
近年来牛顿插值公式被广泛应用于图像插值领域,但是简单的使用牛顿插值公式会造成大量的数据溢出;使用外部存储器进行数据缓冲,占用大量存储资源,使得实时图像延迟一帧。针对以上不足之处设计了一种新的二阶牛顿插值算法在FPGA中的实现方法,首先将牛顿插值公式进行改进,抑制了数据溢出;同时利用FPGA内部逻辑实现数据缓存,在几行时间之内完成数据插值,大大缩短了延迟时间。经实际应用,证明该算法可有效改善边缘模糊和锯齿效应,提高图像质量;同时节省了资源,提高实时性。
Recently Newton＇s formula has been used widely in image interpolation.But using it directly will cause data overflow;what＇s worse,it uses external memory as buffer,not only taking up a lot of resources,but also making images delay a frame.In this paper,a new method to use Second-order Newton interpolation is designed to solve the problems mentioned above.First of all,the Newton＇s formula is modified to eliminate the data overflow.In addition,internal logic unit of FPGA is used,making data interpolation done within several lines.And it can shorten the delay time.The practical applications show that the algorithm can effectively improve the image quality and eliminate the effect of edge blur and mosaics.At the same time it not only save the resources but also improve the systerm＇s real-time effects.
出处
《红外技术》
CSCD
北大核心
2010年第12期723-726,共4页
Infrared Technology
基金
部委预研项目
编号:40405030103
关键词
图像插值
牛顿插值
均差
FPGA
image interpolation
newton interpolation
difference
FPGA
