Self-adaptive robust constrained maximum variance mapping (CMVM) characteristic dimensionality reduction and extraction method for diversified image retrieval of plant leaves

Abstract

The invention discloses a self-adaptive robust constrained maximum variance mapping (CMVM) characteristic dimensionality reduction and extraction method for the diversified image retrieval of plant leaves. On the basis of research on the characteristic extraction of image manifold and selection level, by adoption of a CMVM semi-supervised manifold dimensionality reduction method, the discrimination of positive class local sub-concepts can be kept, and the discrimination of positive and negative classes namely concepts is strengthened. By the invention, a de-noising method and a CMVM strengthening positive local keeping algorithm are provided for keeping the discrimination of the sub-concepts; a linear approximation method is provided for solving the problem of outer point learning of a CMVM sample; an ordered layer maximum interval correlation evaluation function of diversified retrieval is provided for selecting CMVM manifold functions and estimating image intrinsic dimensionality; and a maximum difference intrinsic characteristic method for mining and discriminating positive intra-class sub-concepts from CMVM characteristics is provided for clustering diversified learning, and the diversity of plant image retrieval is improved.

Description

technical field [0001] The invention relates to a feature dimension reduction and selection method, in particular to an adaptive robust CMVM feature dimension reduction and extraction method for diverse image retrieval of plant leaves. Background technique [0002] Plants are one of the life forms with the largest number of species and the widest distribution on the earth. They maintain the balance of carbon dioxide and oxygen in the atmosphere through photosynthesis. At the same time, plants are an important source of food for human beings, as well as resources necessary for human production and life. In addition, plants also play a vital role in soil and water conservation, desert suppression and climate improvement. According to statistics, there are about 400,000 species of plants on earth, of which about 270,000 have been named and recorded by botanists. There are more than 35,000 species of higher plants in my country alone, accounting for about 10.5% of the world's ...

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to propose an adaptive robust CMVM feature dimensionality reduction and extraction method for diversified image retrieval of plant leaves. The method starts from the image manifold feature extraction and selection level, and aims at diverse image retrieval problems. (1) Propose a robust CMVM manifold algorithm to solve the noise problem in image data; (2) Propose a CMVM manifold sample outlier learning method and eigendimension estimation method based on linear approximation; (3) Propose a method based on " Adaptive CMVM manifold parameter selection and eigendimension estimation method for the static evaluation index of the maximum interval correlation in the "ordered" hierarchy; (4) Propose a positive class intra-class "sub-concept" maximum difference eigenfeature selection method to more effectively distinguish "sub-concepts"

Images