A prototype tool for the detection, segmentation, classification and counting of Lego pieces based on the OpenCV artificial vision library is presented. This prototype arises before the need to automate the complex and tedious task of the inventoried one of Lego kits of the MindStorm serie. In the process of detection and segmentation there have been used skills of threshold and the algorithm of Watershed segmentation. For the process of classification and count have been used two different approaches in the securing of the vector of characteristics of the image: BOW and Naive; as well as vector machines support (SVM) for the classification.
CITATION STYLE
Rodríguez-Garrido, A. J., Quesada-Arencibia, A., Rodríguez-Rodríguez, J. C., García, C. R., & Moreno-Díaz, R. (2018). Automatic inventory of multi-part kits using computer vision. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10671 LNCS, pp. 142–149). Springer Verlag. https://doi.org/10.1007/978-3-319-74718-7_17
Mendeley helps you to discover research relevant for your work.