In order to balance the robustness and building complexity better

a local feature description algorithm based on Laplacian is presented. It is analyzed and illustrated that Laplacian not only has good properties to image Euclidian transformation

zoom

and linear brightness changes

but also can characterize the local structure of the image. On the basis of that

a 64 dimensions descriptor is proposed and built with the response of Laplacian of Gaussian. Some measures such as partitioning the feature point neighborhood with its domain orientation as the starting point and normalizing the feature vector are taken to make the descriptor keep invariant with image rotation and luminance changes. Finally

the descriptor is used to match feature points with the absolute distance as similarity measurement. Simulation results indicate that the proposed descriptor performs better in the case of image zoom

rotation

blurring

illumination varying as well as smaller viewpoint changes

and the matching speed is at least 4 times of SIFT. The proposed feature description algorithm is suitable for exigent real-time performance requirement areas and matching the images of structured scene

between which there is the transformation of rotation

zoom

luminance varying

compression or smaller viewpoint changes.