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

Abstract

The invention relates to a land resource analysis model training method and analysis method based on image recognition, and the method comprises the steps: extracting the convolution features of to-be-detected aerial pictures with different downsampling multiples according to a feature extraction model after obtaining the to-be-detected aerial pictures of a to-be-analyzed region; converting each convolution feature into a corresponding multi-dimensional vector according to a vector conversion model; splicing the multi-dimensional vectors to obtain a to-be-tested feature vector, and obtaining a land type label corresponding to a to-be-analyzed area for analyzing land resources based on the land resource analysis model according to the matching degree of the to-be-tested feature vector and the reference feature vector. Based on this, through the convolution features of different downsampling multiples, the fine difference of the aerial photo to be detected can be distinguished, the anti-interference capability of image recognition is improved, and the accuracy of subsequent land type label output is improved.

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 humans and can be used by humans in the foreseeable future. Land resources include both the natural category, that is, the natural attribute of land, and the economic category, that is, the social attribute of land, which is 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 and green development of the rural economy, in the development of agriculture, forestry, waste disposal, breeding improvement, soil improvement, ecological green space, etc., land resources usually need to be ...

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