road network graphics retrieval for GPS art

GPS art aims to create movement tracks with special meaning by taking paths with a specific shape from the road network, adding fun to the movement and encouraging active participation. However, traditional methods rely heavily on manual assistance, resulting in relatively low efficiency. This study presents a new method to automatically retrieve paths with specific shapes by identifying patterns similar to input patterns in a road network. In this method, a one-to-one or one-to-many correspondence relationship is established between the line segment in the input graph and the line segment in the road network by using the invariant spatial relation such as the turning Angle and length ratio and the graph matching technology. This makes it possible to retrieve individual patterns in the road network. Based on this, this paper proposes an algorithm based on greedy strategy to solve the problem of combinatorial graph retrieval. Search results are evaluated to ensure high quality. Simulation and real road network data from 5 different regions are used to verify the accuracy and effectiveness of the proposed method. The research results can automatically generate the path of input graphics and contribute to the promotion of GPS art. In addition, shape constrained graph retrieval expands the application field of spatial scene matching.