Title |
PET plastiko dribsnių rūšiavimo sistemos sukūrimas ir tyrimas / |
Translation of Title |
Development and research of a PET flake sorting system. |
Authors |
Stončius, Vaclovas |
Full Text |
|
Pages |
57 |
Keywords [eng] |
PET flakes ; color spaces ; optical sorting ; MATLAB |
Abstract [eng] |
The main focus of this thesis is to develop a system that is able to distinguish clear white PET plastic flakes from the other colors automatically and with sufficient throughput. Colored flakes are distinguished by using different color spaces and different position search methods are applied to determine each flake's position. The most important parameters for the optical sorting system are the speed and recognition accuracy. The first part of this paper represents the analysis of the literature, how PET flakes are made and why it is important to sort flakes by color, also four most commonly used color spaces are presented: RGB, HSV, L*a*b and YCbCr. Different algorithms for finding an object location are also presented: union-find, flood fill and BLOB analysis. The second part of this paper represents the algorithm which was developed to sort white clear flakes. The algorithm was tested with different color spaces and algorithms for finding an object location. The effectiveness of each color space and algorithms for finding object location were found and the best ones were chosen. |
Dissertation Institution |
Kauno technologijos universitetas. |
Type |
Master thesis |
Language |
Lithuanian |
Publication date |
2017 |