Image fine classification method based on Fisher Vector

The invention provides an image fine classification method based on Fisher Vector. The image fine classification method is used for solving the technical problem of low classification accuracy rate in the existing image fine classification method, and comprises the steps of: reading data of an image library to obtain a training image set and a test image set which contain various classes; extracting RGB features of each image in the training image set and the test image set; solving mixed Gaussian parameters described by the RGB features of the training image set; solving Fisher Vector eigenvectors of a matched image block set; solving final feature descriptions of the training image set and final feature descriptions of the test image set; utilizing SVM for training the final feature descriptions of the training image set to obtain a classification model; and utilizing the classification model for classifying the final feature descriptions of the test image set. The image fine classification method has the a