Prior to 2010, different image extraction projects were treated in the field of image recognition by mixing handmade features, manually created image local features and machine learning techniques. However, additional learned in the classroom image recognition methods have been presented since the year 2010. In general object recognition contests, deep learning-based image recognition strategies have beat earlier technologies by a big amount. In order to better understand how deep learning is used in the field of image identification, this conference will also discuss the latest progress in deep learning-based autonomous driving.