The invention provides a location recognition and relative positioning method and a location recognition and relative positioning system adapting to changes of visual characteristics. The location recognition and relative positioning method comprises the steps of: acquiring an outdoor scene image; acquiring a scene recognition result based on the outdoor scene image, a pre-constructed scene library and a scene classifier; and acquiring a relative positioning result based on the scene recognition result and a pre-constructed environment profile map. The location recognition and relative positioning method and the location recognition and relative positioning system adopt the machine learning technology to improve the location recognition accuracy, and extract the image features by means ofa deep convolutional neural network model trained on a large-scale labeled image data set instead of artificially selecting a fixed feature extraction operator, therefore, the feature extraction can better adapt to changes of the visual features.