A survey of 2D shape representation: Methods, evaluations, and future research directions

摘要: Abstract In the past few years, research studies in image-based shape representation have been proliferating due to its usefulness and importance for various application. This field has evolved, from simple descriptor-based instance retrieval utilization of machine learning approaches. Thus, this papers aims provide a comprehensive survey summarize overall view topic. It covers several concepts including traditional descriptors, boundary region partitioning strategies, more advanced techniques which commonly exist recent studies. manuscript discusses advantages drawbacks these methods by providing comparisons evaluation results on well-known public datasets under types similarity metrics assessment procedures. To complete survey, it also suggests diverse possibilities future directions.