Image recognition method and corresponding device, storage medium and electronic equipment

Abstract

The invention provides an image recognition method and device, a storage medium and electronic equipment. The image recognition method comprises the steps of obtaining a first image and a second imageof a target area; wherein the acquisition conditions of the first image and the second image are different; determining a difference image based on the heterogeneity information of the first image and the second image; carrying out FCM calculation and secondary classification on the difference image to obtain an initial label field image; and identifying changes in the second image relative to the first image based on the difference image and the initial label field image. The edge continuity and the region integrity of the obtained difference image are good, the probability that when an ICMRF algorithm is used for calculating the difference image, iteration frequency is increased, and the difference image is prone to falling into the local optimal solution is reduced, the change of the second image compared with the first image can be accurately recognized, and then the accuracy of the change region is improved.

Description

technical field [0001] The present disclosure relates to the technical field of radar image processing, and in particular, to an image recognition method, a corresponding device, a storage medium, and an electronic device. Background technique [0002] Synthetic Aperture Radar (SAR) has the advantages of all-day, all-weather, and large-scale observation, and can also obtain high-spatial-temporal resolution and high-precision SAR images. It has broad application prospects in the field of disaster change detection. Among them, SAR image change detection refers to the analysis and comparison of multiple SAR images obtained in the same area at different times and under different imaging conditions, and then determine the changed area. Point change information is more helpful to the understanding and prediction of disasters; therefore, change detection in SAR images has been specifically applied in urban monitoring, land use, and disaster early warning. [0003] In the prior art...

AI Technical Summary

Problems solved by technology

[0003] In the prior art, during the change detection of SAR images, that is, the process of using SAR images to determine the change area in the target area, there are the following problems: 1. Only using SAR images to directly synthesize a difference map, which is affected by speckle The impact of noise is serious, and the edges are discontinuous, and the integrity of the region is poor, which leads to low accuracy of the change region determined by using the difference map; 2. Directly using the FCM algorithm for secondary classification, resulting in many false inspection points, which also leads to The accuracy of the determined change area is low; 3. When using the ICM-MRF algorithm to calculate the difference map, as the number of iterations increases, it is easy to fall into a local optimal solution, and due to the Markov property, it is easy to cause false detection points, thereby reducing the certainty. The accuracy of the change area within the target area

Images