Image content identification method and apparatus

Abstract

The invention discloses an image content identification method and apparatus. The method comprises: extracting a CNN feature of a to-be-identified image; inputting the to-be-identified image into a pre-established image classifier set, obtaining a probability value output by each image classifier in the image classifier set according to the CNN feature of the to-be-identified image, and selecting at least one image classifier as a target image classifier from the image classifier set according to the probability value output by each image classifier; determining the similarity between each image in the target image classifier and the to-be-identified image, and selecting multiple images as target images from the target image classifier according to the similarity; and marking the to-be-identified image by using a type tag corresponding to a class cluster in the target image classifier and an entity tag corresponding to the target image in the target image classifier. With the adoption of embodiments of the invention, the accuracy of image identification can be improved.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to an image content recognition method and device. Background technique [0002] With the development of Internet technology and social media, users need to process more and more pictures. There are a large number of pictures uploaded by users on the Internet and social networking software. Most of these pictures do not have content information marked, such as: geographic location pictures, famous Pictures of scenic spots or buildings, etc. make it impossible for users to determine what these pictures are or where they are. Therefore, a method for identifying these pictures is needed. [0003] In the prior art solution, first extract the features of the image to be identified, then retrieve some feature data most similar to the features of the image to be identified in the image database, and finally use the relevant web page text of the similar data to infer the co...

AI Technical Summary

Problems solved by technology

However, this method has the following disadvantages: first: it relies on the webpage text data related to the image to be recognized, but the content labels extracted from the text data contain more noise; second: the retrieved image database is large in scale, including More noise, it is easy to retrieve pictures with similar content but different semantics, resulting in the final label is not accurate enough; third: the range of recognized picture content is large, not accurate enough

Images