Crop phenotypic parameter extraction method and system based on multi-view vision

Abstract

The invention discloses a crop phenotypic parameter extraction method and system based on multi-view vision. The method comprises steps of reconstructing to obtain a three-dimensional point cloud of acrop group through an MVS method; reducing the number of points in the three-dimensional point cloud and removing outliers through down-sampling filtering and a statistical denoising preprocessing method; removing ground point clouds through RANSAC, extracting plant height information by referring to the ground, extracting point clouds of plant individuals and a single leaf through an Euclidean clustering algorithm and a region growing algorithm, and finally performing point cloud surface reconstruction and calculating leaf length and leaf width phenotypic parameters through a Delaunay triangulation algorithm. According to the method, only one digital camera is needed, compared with a binocular stereo vision method, the calculation cost for reconstructing the accurate and dense three-dimensional point cloud is lower, and a complex calibration process is not needed; the reconstruction effect is good, and texture and color information of real crops is achieved; non-contact nondestructive measurement of phenotypic parameters of the crop plants is carried out; and a feasible scheme is provided for providing a rapid reconstruction three-dimensional model for the MVS and measuring plantphenotypic parameters at low cost.

Description

technical field [0001] The invention relates to the extraction of crop phenotype parameters, in particular to a method and system for extracting crop phenotype parameters based on multi-eye vision. Background technique [0002] In the field of agriculture, in order to cope with the development of plant genomics analysis technology, the development of plant phenotype analysis technology is urgently needed. Breeding scientists need to combine gene sequence information and corresponding crop phenotype information to screen out high-yield, high-quality high-quality genes, so as to meet the increasing demand for future food production. Therefore, lagging phenotypic measurement techniques will restrict breeding scientists from digging out genetic information that regulates crop traits. High-throughput, high-precision, nondestructive and nondestructive phenotyping measurements of plants have become key issues in the field of agriculture. Researchers have made many explorations on...

AI Technical Summary

Problems solved by technology

However, the effect of stereo vision reconstruction generally requires a higher computational cost, is more sensitive to environmental lighting, and is also affected by the occlusion of plant organs themselves.

Images