摘要
本文着重探讨经预处理和区域分割后的单幅图像如何识别出颗粒的方法。首先将区域预截取图像作差分运算并进行垂直投影定位,根据投影量对图像再作截取,然后将截取到的图像块进行阈值分割以得到颗粒所在区域,并通过计算求得颗粒中心点位置。
Trajectory and velocity of saltating particles are vital to mechanics of sediment transport. Computer image processing and recognition techniques can be used to acquire high precision data on saltation of particles. The images were taken from the experiments on the bed-surface of nonuniform sediment. In this paper, the emphasis is put on saltating particle recognition, which is performed on a single pre-processed and segmented image. Difference operation on the segmented image is performed before the vertical projective positioning. Then the image is further segmented by an empirical threshold to find the particle and its central position is estimated.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2000年第3期357-360,共4页
Pattern Recognition and Artificial Intelligence
基金
国家自然科学基金
关键词
图像识别
颗粒跳跃
非均匀沙
计算机
Image Recognition, Particle Saltation, Nonuniform Sediment
