| 摘要 |
Image recognition is performed based on a surrounding image and a recognition template used for the image recognition of a marker object, and a recognition confidence level used for determining if the marker object can be recognized in the surrounding image is calculated. A determination is made if the recognition confidence level has increased as compared with the recognition confidence level calculated based on the surrounding image acquired at the guidance output point. If it is determined that the recognition confidence level has increased, the image of the marker object, generated based on the surrounding image acquired at the guidance output point, is stored as a new template to be used for the image recognition of the marker object. This increases the possibility to recognize the marker object based on the new template, thus increasing the recognition accuracy of the marker object. |
