Research On Large Scale Images Recognition From Internet

With the rapid development of deep learning,computer vision has been wildly applied in traffic,medical,security,advertising,personal image video analysis and other fields.In computer vision image recognition played a key role.Image recognition based on deep learning requires a large amount of labeled data for model training,while manual collection of labeled data requires a lot of money and time.At the same time,with the development of image search engine and personal image sharing website,a large amount of image data can be acquired from the Internet,which has a large number of noise and is difficult to be directly used for training.Based on the current problems,this paper proposed a method of training image recognition model using the massive data on the Internet:using Web crawlers to automatically crawl the image data from the Internet;automatically cleaning data according to the deep visual features of the images;efficiently train the models on a pre-trained model.In order to deal with the massive data,this paper makes use of the cloud computing platform to build the service on the distributed cluster,which accelerates the whole process.Finally,based on the research of this paper,the prototype system of image recognition based on Internet and cloud computing is implemented,and the effectiveness of the method has been verified by experiments.