Comparative Study of Thresholding Techniques to Evaluate Cork Quality

B. Paniagua-Paniagua, M.A. Vega-Rodríguez, J.A. Gómez-Pulido, and J.M. Sánchez-Pérez (Spain)


Automatic thresholding, image processing, local thresholding, global thresholding, cork quality.


The most important use of cork is the development of stoppers/disks in the different kinds of liquor industries. The main problem of cork is that it is a natural material. This means that it is a heterogeneous material, and therefore, it is not possible to obtain two or more equal cork stoppers/disks. This has always been a serious problem in order to do an automatic classification of cork stoppers/disks based on their quality (usually, seven different quality classes exist). The quality of cork stoppers/disks is mainly based on their defects. These defects can be obtained using thresholding techniques. In this paper we attempt to evaluate the performance of a great number of automatic thresholding methods in order to conclude which of these methods is the most appropriate one for the cork industry. This paper evaluates a total of 13 different thresholding techniques, including both global and local thresholding methods. In conclusion, after the evaluation of all these techniques, for the cork industry, the moment-preserving method is the global thresholding technique which obtains the best results, and the statistical thresholding method is the local thresholding technique that produces the best results.

Important Links:

Go Back