Land resource analysis model training method and analysis method based on image recognition

Abstract

The present invention relates to a land resources analysis model training method and analysis method based on image recognition. After obtaining the aerial photographs to be tested in the area to be analyzed, the convolution features of the aerial photographs to be tested with different downsampling multiples are extracted according to the feature extraction model. And convert each convolutional feature into a corresponding multi-dimensional vector according to the vector conversion model. Then, the multi-dimensional vectors are concatenated to obtain the feature vector to be measured, and based on the land resource analysis model, according to the matching degree between the feature vector to be measured and the reference feature vector, the land type label corresponding to the area to be analyzed for analyzing land resources is obtained. Based on this, the convolution features of different downsampling multiples are beneficial to distinguish the subtle differences of the aerial images to be tested, improve the anti-interference ability of image recognition, and improve the accuracy of subsequent land type label output.

Description

technical field [0001] The invention relates to the technical field of land utilization, in particular to an image recognition-based land resource analysis model training method and analysis method. Background technique [0002] Land resources refer to the land that has been used by human beings and can be used by human beings in the foreseeable future. Land resources include both the natural category, that is, the natural attributes of the land, and the economic category, that is, the social attributes of the land, which are the means of production and labor objects of human beings. Especially in the rural economy, the rational use of land resources is often one of the basic foundations of agricultural production, forestry production or aquaculture production. In the process of sustainable green development of the rural economy, in the development of agriculture, forestry, waste treatment, breeding improvement, soil transformation, and ecological green space, it is usually...

AI Technical Summary

Problems solved by technology

However, with the development of urbanization, the reduction of the rural production population, the gradual concentration and scale of production based on land resources in the countryside, the traditional manual observation method will undoubtedly consume a lot of valuable human resources, or face a shortage of human resources The problem

At the same time, the manual observation method also has the problems of low efficiency and low accuracy, and it is difficult to meet the centralized and large-scale needs of rural production based on land resources.

Images