Research On The Recognition Of The Number Of The Grain Bags Based On Watershed Algorithm And Chain Code

The supervision and audit is the core of our country's grain reserve management, but there are many reasons leading to the inefficient manage, and the relevant departments couldn't understand the inventory level of inventory level grain accurately and timely, and then, the ability of the macroeconomic regulation through food is insufficient.As a new discipline, digital image processing technology develop in a few decades of time rapidly, and widely used in many fields, such as military, industry ,aerospace and medical science. Image segmentation is the most basic link of the computer image analysis and understanding, and the segmentation result is related to the performance of the subsequent algorithm directly. This paper firstly introduces the definition and the methods of the image segmentation technology, mainly the color image segmentation. And in order to make the segmentation effect better, this paper uses the algorithm based on the enhancement of the S and I color components before the edge detection.Based on the analysis of the differences, the advantages and disadvantages between the classical operators, this paper introduces mathematical morphology and the basic principle of operators, the theory and the processing steps of the watershed algorithm. Through the experiment show the advantages and disadvantages of the algorithm of traditional watershed image edge detection, and improve the watershed algorithm by using the mathematical morphology top-hat algorithm and bot-hat algorithm.The paper introduces the basic theory of the Freeman chain code, the vertex chain code and the crack code, discusses the mapping relationships between them .And then, presents identification algorithm based on chain code number. Theoretically, there are some good effects of the algorithm for grain heap numbers recognition.