首页>
外国专利>
METHOD OF IMPROVING CLASSIFICATION ACCURACY OF SNS IMAGE DATA FOR TOURISM USING SPACE INFORMATION DEEP LEARNING TECHNOLOGY, RECORDING MEDIUM AND APPARATUS FOR PERFORMING SAME
METHOD OF IMPROVING CLASSIFICATION ACCURACY OF SNS IMAGE DATA FOR TOURISM USING SPACE INFORMATION DEEP LEARNING TECHNOLOGY, RECORDING MEDIUM AND APPARATUS FOR PERFORMING SAME
A method of improving the classification accuracy of SNS image data for tourism using a space information deep learning technology includes the steps of: acquiring geo-tagged image data posted on an SNS, and classifying the image data according to an image-only classification system for tourism purposes through a convolutional neural network (CNN) based on tourist activity information (What); extracting a cluster for each category detailed item through a density-based spatial clustering of applications with noise (DBSCAN), by using category information labeled on the classified image data; comparing location information (Where) extracted from the cluster with the labeled value of the activity information (What), and updating the activity information (What) on the basis of the location (Where) information upon mismatch; reclassifying the image data according to the image-only classification system for tourism purposes through the convolutional neural network (CNN) based on the updated activity information (What); and measuring the accuracy of the image data classification through overlap analysis of the updated activity information (What) and an actual location. Accordingly, the location information is checked through DBSCAN of the category-labeled photo data, and the location information is updated to improve the classification accuracy of the images for tourism purposes when there is an error.;COPYRIGHT KIPO 2020
展开▼