Multi-dimensional geographic scene identification method fusing geographic region knowledge

Abstract

The invention discloses a multi-dimensional geographic scene identification method fusing geographic region knowledge. The method comprises the steps of preprocessing images in a database to obtain satisfied geographic scene images; obtaining object region image blocks by utilizing a method for quickly searching for object regions in the images; pre-training the obtained object region image blocks of the geographic images by using a deep convolutional neural network, performing an accurate adjustment process until the performance of the deep convolutional neural network of the scene images is no longer improved, and fusing feature matrixes into output eigenvectors; pre-establishing a geographic entity noun keyword dictionary by acquired entity noun data in geographic scene classification, performing word segmentation on target identification result data to obtain key words in a target identification result, and establishing text features; and fusing the text features and multi-dimensional image features into eigenvectors as inputs, realizing cross-media-data identification classification, and realizing scene classification fusing geographic entity information.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to multi-dimensional geographic scene recognition technology. Background technique [0002] Scene classification, that is, to complete the automatic recognition of image scene categories (such as mountains, forests, bedrooms, living rooms, etc.) based on the features contained in the scene image, is an important branch in the field of image understanding, and has become an important branch of multimedia information management, computer vision, etc. The hot issue in fields such as, is subjected to the extensive attention of researcher. Scene classification is of great significance to the development of multimedia information retrieval and other fields, and has broad application prospects and theoretical significance in many fields. [0003] With the advent of the big data era, deep convolutional neural networks with more hidden layers have more complex network structures, a...

AI Technical Summary

Problems solved by technology

[0007] When CNN extracts deep-level features of images, it builds a multi-layer network structure, which requires a large number of labeled data samples to train network parameters. However, the cost of manual labeling in actual scene images is high, resulting in the fact that data samples are often used in scene classification. The amount is small, and the classification accuracy of the traditional CNN method is not high

Images