Non-Periodic Translation Symmetry Detection Using Global Self Similarity

摘要: This paper addresses a specific example of nonperiodic translation symmetry and presents an algorithm to automatically detect multiple poles, or their shadows, in aerial imagery by looking for consistent overlapping regions self-similarity across non-urban scene. The does not rely on having pole template knowing its exact size. For each image patch, similar (or blobs) are found the whole using normalised cross-correlation. blobs filtered based known orientation approximate then clusters together all patches that have large number mutually blobs, indicating common degree between them. scenes, these likely identify shaped poles. Results actual demonstrate algorithm's potential most poles with few false alarms, shows superior performance template-matching approach.