Image Feature Extraction And Classification Research Based On Ensemble Learning Model

With the development of the internet,data including images are growing rapidly.The processing of these massive images has become a considerable challenge in the new era.In addition,various images have different sizes and the images are often sensitive to the interference of the lights and shadows,which has brought a great challenge to the image feature extraction and classfication.How to optimize image classification to learn more information has attracted much attention recently.In this article,the image feature extraction and classification algorithm based on the ensemble learning model is proposed.Specifically,three representative image feature extraction methods are combined via the ensemble learning model,including onvolution neural network,scale-invariant feature extraction and histogram of gradient histogram.Therefor the accuracy of classification is expected to be improved.At the same time,we also incorporate the transfer learning into the convolution neural network,which can overcome the inaccurate classification arising from the lack of training samples.In order to verify the performance of our proposed method,a large number of the urban landscape photos were collected from the web sites.We first classify each photos accroding to the The Air Quality Index.Then these photes construct database to validate the performance of our proposed method.The simulation results show that our proposed method is superior to separated classification algorithms...