基于实时性的图像插值加速算法研究

Interpolation accelerated algorithm research based on real-time image

插值算法在图像处理中扮演着重要角色,本文介绍了双线性插值的原理,分析了浮点运算对算法效率的制约,提出了通过把像素值放大一定倍数转换浮点运算为移位运算的算法加速方案,重点设计了插值加速算法,然后对加速算法的运行效率和图像质量进行了对比测试,证实了加速算法在保证图像质量的基础上提高了插值效率,可以在实时系统中广泛应用。

Interpolation algorithm plays an important role in image processing. The article de- scribes the principle of bilinear interpolation, analysis of floating-point arithmetic constraints on the efficiency of the algorithm proposed by the pixel value convert floating-point operations to enlarge a certain multiple of the shift operator algorithm to accelerate the program, focused on the design to accelerate the interpolation algorithm. Then comparison test about accelerate the algorithm efficiency and image quality is conducted. It is confirmed that the accelerated al- gorithm can improve interpolation efficiency and ensure image quality. It can be widely used in real-time system.