一种嵌入式颗粒显微图像统计系统的设计与实现

为了对颗粒显微图像进行有效地统计与识别,设计了一种以DSP为核心的嵌入式图像采集处理系统.模拟视频信号由视频A/D转换为数字图像,DSP首先利用中值滤波、自适应阈值分割等算法对图像进行预处理,将目标从背景中分离出来,然后采用扫描标记的算法对目标区域加以标记,最后实现对颗粒的数目及面积等几何尺寸的统计.实验结果表明,系统不仅能够对颗粒进行全面、客观地统计分析,而且简单、方便,具有重要的实用价值.

Design and Realization of an Embeded Granule Micrograph Statistics System

In order to achieve effectively statistics and recognition of granule micrograph, an embedded image gathering and processing system that take a DSP as the core processor is put forward. The digital image can be obtained by a video A/D. By using median filtering and auto-adapted thresholding, the DSP preprocessed the granule micrograph firstly, separating the objects from the background,then, marked the objects fields through the scanning mark algorithm and finally realized the statistics of the granule number and the geometry parameters such as area etc. The experimental result indicated that, this system not only can carry on the statistical analysis to the granule comprehensively and objectively, moreover, it is simple and convenient, has the important practical value.