Optimal scale selection method and device and computer readable storage medium

Abstract

Embodiments of the invention disclose an optimal scale selection method and device, and a computer readable storage medium. The method comprises the steps of generating the segmentation result graphsunder different scale parameters based on an input remote sensing image; selecting a training sample from the generated segmentation result graphs, training the training sample to obtain a multi-scaleclassification model, then carrying out multi-scale classification on the remote sensing image to be classified, and generating the classification result graphs under different scale parameters; selecting a plurality of target classification result graphs from the classification result graphs, and respectively calculating the classification precision associated with each target classification result graph; and carrying out optimal scale selection after classification precision statistics. By implementing the invention, the multi-scale classification result is used for precision evaluation, the optimal scale is selected according to the precision evaluation statistical result, the applicability is wider, the analysis requirements under more scenes can be met, the conversion from the optimal segmentation scale to the optimal classification scale is achieved, and the defects that the efficiency is low and the effect is poor when the analysis scale is selected at the segmentation stage are overcome.

Description

technical field [0001] The invention relates to the field of remote sensing image processing, in particular to an optimal scale selection method, device and computer-readable storage medium. Background technique [0002] Scale is the window for human beings to understand things. In earth observation, due to the complexity and variety of surface types, different scales of land objects coexist, and the same area also shows different attributes when observed from different scales. Therefore, the issue of scale is a matter of earth observation It is one of the main challenges and also one of the key issues of the Object-Based Image Analysis (OBIA) method. In many applications, the scale analyzed by the OBIA method is embodied as the scale parameter of the segmentation algorithm. Scale, as the core parameter of the segmentation algorithm, determines the scale of the obtained analysis primitives, and choosing an appropriate segmentation scale is the primary problem faced by many ...

AI Technical Summary

Problems solved by technology

[0004] The main purpose of the embodiments of the present invention is to provide an optimal scale selection method, device, and computer-readable storage medium, which can at least solve the problem of obtaining multiple scale analysis results when performing OBIA scale selection in the related art. Selecting a single optimal scale can only meet the needs of some specific application scenarios, and the applicability is relatively limited

Images