Image Processing of Layer Cutting of Compacted Root-soil of Mung Bean

摘要

The important basis for three-dimensional simulation of plant root growth and root architecture is to accurately obtain the parameters of root architecture in the process of root growth. However, the growth of roots in the soil is difficult to observe. Study for finding the effective method to obtain root architecture parameters has been the difficulty. Based on the assumptions of mung bean root growth in different compacted soil, this paper designed a method to layer cut the mung bean roots which grew in soil. The experiment acquired the root-soil image with digital camera, then identified the root in soil-root fault image by image processing, and got the coordinates of the root system. The root-soil samples were cut in the vertical direction in the mung bean cultivation container and cutting sectional images were collected. Gray-scale transformation, noise filtering, image binaryzation, and threshold segmentation processing were respectively carried on the root-soil sectional images by using image processing in MATLAB program and the root position coordinates were got in the root-soil fault image. Because it was difficult to recognize the root from root-soil fault images in different soil compaction which would affect the mung bean root architecture, this paper used different filter to eliminate the sample images of the grayscale disturbance and obtained the root distribution.