| 摘要 |
Interest points are markers anchored to a specific position in a digital image of an object. They are mathematically extracted in such a way that, in another image of the object, they will appear in the same position on the object, even though the object may be presented at a different position in the image, a different orientation, a different distance or under different lighting conditions. The goal is to match interest points in one image with corresponding interest points in another image. Typically, this involves the construction of a descriptor, which is both computationally expensive and resource-intensive. The described invention matches interest points without the construction of conventional descriptors and uses spatial coherency information to increase the accuracy of the match. The process uses a dataset for each image to be matched which is generated from an initial dataset of interest points in each image which includes information identifying position, scale and orientation of each interest point. From these datasets subsequent datasets of objects are calculated wherein each object relates a base interest point to a the normalized position and orientation of a satellite interest point or points. Objects from these databases are compared to identify matches from which the related interest points may be determined and matched. The base interest objects and satellite interest objects are linked in the database by objects termed worms. |
