Shape similarity measure using turn angle cross-correlation for oyster quality evaluation

An overview of the oyster industry in the US shows that oyster grading occurs at harvest, wholesale, and processing markets. Currently whole oysters, also called shellstock, are graded manually by screening and sorting based on diameter or weight, which does not concern shape appearance. However, studies show that customers prefer oysters that are round and 2 in. in diameter. This paper presents a simple and accurate shape analysis method for size and shape-based oyster grading. The proposed algorithm is currently programmed to perform shape analysis to grade oysters into good quality, banana, and irregular grades and can be extended to accommodate other grades acceptable for market. Other shape categories can be added with a simple configuration procedure. The contour of the oyster is extracted and its data points are reduced to 50 points using either curve evolution and fixed interval sub-sampling data reduction methods. A simple and accurate shape similarity measure method called turn angle cross-correlation (TAC) is developed to categorize oyster shell shape and other similar applications. Experimental result shows that this method performs well and is suitable for commercial use.