Two-dimensional pattern matching with k mismatches

作者:

Highlights:

摘要

We give an algorithm which finds all occurrences of an m1 × m2 pattern array embedded as subarrays in an n1 × n2 array of text, where at most k mismatches are allowed per occurrence. The algorithm runs in time O((k + a)(b log b + n1n2)), where a = min(m1, m2) and b = max(m1, m2). This improves upon the previously best known algorithm, and is asymptotically optimal for k ≈ a.

论文关键词:Pattern matching,Computer vision

论文评审过程:Received 20 September 1989, Accepted 29 January 1990, Available online 19 May 2003.

论文官网地址:https://doi.org/10.1016/0031-3203(91)90114-K