Soil image segmentation and extraction method based on super-green index and super-pixel

Abstract

The invention relates to a soil image segmentation and extraction method based on a super-green index and a super-pixel, and the method comprises the following steps: firstly segmenting and removing the most common interference element green vegetation in a soil image through the super-green index; then, using a green vegetation area as a mask to be overlapped with an original image, reducing interference of non-soil elements on subsequent segmentation, and enhancing the stability of the whole algorithm; and finally, performing superpixel segmentation on the image after mask superposition, merging superpixel blocks based on the similarity corresponding to the Euclidean distance, and removing other interference elements in the soil image such as straws and small block shadows through segmentation. According to the whole design scheme, the soil elements in the soil image are automatically and quickly segmented and extracted, and the defects that time and labor are consumed when non-soil elements are removed from the soil image through a manual method are overcome. In addition, the method overcomes the defects that an existing automatic segmentation method can only extract part of soil areas in the image and is not accurate enough, and all soil element areas in the image can be extracted more accurately.

Description

technical field [0001] The invention relates to a method for segmenting and extracting soil images based on super green index and super pixels, and belongs to the technical field of soil image processing. Background technique [0002] Soil photos taken in the field are an important source of soil color and other attributes. However, soil photos taken often contain non-soil elements such as green plants and straws. These non-soil elements often have the characteristics of small area, large number, and scattered distribution. Non-soil elements will bring large errors to the acquisition of soil color and other attributes, and it often takes a lot of time and effort to remove them from the image by manual methods. Therefore, the acquisition of soil elements in existing images is very inefficient. Contents of the invention [0003] The technical problem to be solved by the present invention is to provide a soil image segmentation and extraction method based on the super green ...

AI Technical Summary

Problems solved by technology

[0002] Soil photos taken in the field are an important source of soil color and other attributes. However, soil photos taken often contain non-soil elements such as green plants and straws. These non-soil elements often have the characteristics of small area, large number, and scattered distribution. Non-soil elements will bring large errors in the acquisition of soil color and other attributes, and it often takes a lot of time and effort to remove them from the image by manual methods

Therefore, the acquisition of soil elements in existing images is very inefficient

Images