Landmark recognition on iPhone was proposed in this paper,so as to enable the user to take photos of some famous landmarks in the world and return the result of what this landmark is as well as provide some useful inf...Landmark recognition on iPhone was proposed in this paper,so as to enable the user to take photos of some famous landmarks in the world and return the result of what this landmark is as well as provide some useful information about this landmark such as travel guide,history and interesting activities.The approach to recognize the landmark is by extracting features,using bag-of-words model and support vector machine to train the dataset,hence to get the classification model.Then this model is used to predict photos the user has taken.This article compares the different algorithm like SIFT,Gist and Hog to extract features and build a classification model.This article uses PC as the server to recognize landmark photos and return results to users.展开更多
基金Project supported by the National Natural Science Foundation of China(Grant No.13BF053)
文摘Landmark recognition on iPhone was proposed in this paper,so as to enable the user to take photos of some famous landmarks in the world and return the result of what this landmark is as well as provide some useful information about this landmark such as travel guide,history and interesting activities.The approach to recognize the landmark is by extracting features,using bag-of-words model and support vector machine to train the dataset,hence to get the classification model.Then this model is used to predict photos the user has taken.This article compares the different algorithm like SIFT,Gist and Hog to extract features and build a classification model.This article uses PC as the server to recognize landmark photos and return results to users.