UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Template-based sketch recognition using Hidden Markov Models Flor, Roey


Sketch recognition is the process by which the objects in a hand-drawn diagram can be recognized and identified. We provide a method to recognize objects in sketches by casting the problem in terms of searching for known 2D template shapes in the sketch. The template is defined as an ordered polyline and the recognition requires searching for a similarly-shaped sequential path through the line segments that comprise the sketch. The search for the best-matching path can be modeled using a Hidden Markov Model (HMM). We use an efficient dynamic programming method to evaluate the HMM with further optimizations based on the use of hand-drawn sketches. The technique we developed can cope with several issues that are common to sketches such as small gaps and branching. We allow for objects with either open or closed boundaries by allowing backtracking over the templates. We demonstrate the algorithm for a variety of templates and scanned drawings. We show that a likelihood score produced by the results can provide a meaningful measure of similarity to a template. An example-based method is presented for setting a meaningful recognition threshold, which can allow further refinement of results when that template is used again. Limitations of the algorithm and directions for future work are discussed.

Item Media

Item Citations and Data


Attribution-NonCommercial-NoDerivatives 4.0 International