Learnable difference algorithm-based remote sensing image change detection method

Abstract

The invention provides a learnable difference algorithm-based remote sensing image change detection method. The method comprises the following steps: firstly, executing variation and crossover operations in each iteration process of a differential evolution algorithm; secondly, randomly selecting a part of individuals from the original population and the crossed population to enter a new population, selecting individuals from the new population and marking the individuals to obtain training samples; then, training the neural network, and selecting individuals entering the next generation of population by using the trained neural network; and performing iteration in this way to obtain a population approaching a real optimal value, calculating an optimal fuzzy relation matrix, allocating each pixel in the difference graph to the category of the maximum fuzzy value, and finally completing change detection. The method provided by the invention has good optimization capability and convergence rate, and is high in detection precision.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a remote sensing image change detection method based on a learnable difference algorithm. Background technique [0002] Change detection for remote sensing images has a wide range of application scenarios in reality. For example, in the assessment and processing of natural disasters, the use of remote sensing image change detection can quickly analyze the disaster-affected area, so as to implement effective rescue and avoid danger. [0003] Traditional remote sensing image change detection algorithms are mainly divided into four categories: threshold method, clustering method, graph cut method and level set method. Among them, the more commonly used algorithms are the threshold method and the clustering method, while the graph cut method and the level set method are usually used in the initialization part of the first two algorithms due to the problem of the ...

AI Technical Summary

Problems solved by technology

[0004] However, clustering algorithms usually have problems such as slow convergence, optimization methods that cannot reach the ideal optimal value, and excessive computational complexity, which is not conducive to the practical application of the algorithm.

Images