An Image Processing Method for Grains Counting

In the process of quality analysis of grains, number counting is one of the key steps for agricultural production. After the introduction of the traditional counting methods, an approach based on computer version and image processing technique is proposed in this paper. The overall structure including hardware and software of system is introduced firstly. Then, the image processing methods are described in details, which are mainly composed of gray transformation, denoising, binary conversion, mathematical morphology analysis, et al.. Finally, an experiment is conducted to test the validity of the algorithm, and the result shows that this method is effective and can realize the automatic grain counting.