Remote sensing target segmentation and automatic splicing method based on voting mechanism

Abstract

The invention relates to the technical field of remote sensing, and discloses a remote sensing target segmentation and automatic splicing method based on a voting mechanism, and the method comprises the following steps: S1, segmenting an image in the form of a sliding window, and ensuring that each pixel appears in nine mutually overlapped blocks at the same time; S2, recording coordinate information of a central point of each block, and matching a memory address for each block; and S3, distributing the different blocks to a plurality of threads or distributed operation programs of the interpretation model for parallel operation, and outputting the category of the pixel and the corresponding estimation probability at the same time. According to the remote sensing target segmentation and automatic splicing method based on the voting mechanism, the segmentation result considers the forms of the image pixels at the central position and different edge positions, the edge effect occurring in block calculation is effectively avoided, and the technology can be applied to remote sensing image target detection and boundary extraction, and be used for designing a large remote sensing image block calculation and parallel processing program.

Description

technical field [0001] The invention relates to the technical field of remote sensing, in particular to a method for segmenting and automatically combining remote sensing targets based on a voting mechanism. Background technique [0002] Remote sensing refers to non-contact, long-distance detection technology, generally refers to the use of sensors / remote sensors to detect the radiation and reflection characteristics of electromagnetic waves of objects. Remote sensing is to detect the target ground object under the condition of being far away from the target and non-contact target objects through instruments such as remote sensors that are sensitive to electromagnetic waves. With the rapid development of remote sensing technology, the spatial resolution of optical remote sensing images is getting higher and higher. Due to the amount of remote sensing data Large or model input limitations, when using remote sensing images for ground object segmentation, the image must be di...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the deficiencies of the prior art, the present invention provides a remote sensing target segmentation and automatic stitching method based on a voting mechanism, which can avoid edge effects in the process of segmenting remote sensing objects with a fixed window, and realize large-scale remote sensing image block parallel segmentation and the automatic seamless splicing of the output boundary, etc., which solve the problems raised in the above-mentioned background technology

Images