A feature image category information generation method and device

Abstract

The embodiment of the invention provides a feature image category information generation method and device. The method comprises the steps: obtaining an original image; Searching a target area from the original image; Obtaining a feature information image of the original image; Extracting target feature information of the feature information image from the target area; And generating category information of the target area according to the target feature information. By the adoption of the mode, the target area used for displaying the target image in the original image is determined; and thentarget feature information is extracted from the feature information image of the target area to generate corresponding category information, so that the generation accuracy rate of the feature imagecategory information of the original image with the complex background is improved, and the accuracy rate of pushing related information to the user is improved.

Description

technical field [0001] The present invention relates to the technical field of generating feature image category information, in particular to a method for generating feature image category information and a feature image category information generating device. Background technique [0002] With the development of science and technology, computer technology has affected every aspect of people's life. For example, users can quickly browse current hot news or information that users care about through the Internet. In the process of users browsing information, by analyzing the user's browsing content, and formulating personalized information push to users according to the analysis results, it can greatly save the time spent by users searching for information, thereby improving the user's browsing information. For example, when a user browses a picture on the Internet, the element in the picture can be analyzed, and information related to the element can be pushed to the user. ...

AI Technical Summary

Problems solved by technology

[0003] When the user browses the picture, if the elements in the picture are game content, it is impossible to accurately push the information related to the game content to the user, because in the prior art, the way to identify the elements in the image is Directly use machine learning or deep learning-based classification technology to classify the elements in the picture. In practical applications, when game content exists as an element in a picture, it is not necessarily the only element contained in the picture, except The game content itself may also contain a more complex background. If the above method is used, the game content in the picture cannot be accurately identified, and it is impossible to push relevant information to the user for the game content.

Images