| Abstract [eng] |
In the last decades, computer vision technology and image processing have received a lot of attention in the various scientific and industrial fields. One of the most widely used areas of image processing is the information identification. The information identification has become increasingly important in order to ensure product traceability and information accuracy while the industry and automation level are growing. Therefore, this paper explores the task of the information identification in the industry. The first part briefly reviews the computer vision and the character identification. It also overviews the k–Nearest Neighbors KNN, the Support Vector Machines SVM, the Artificial Neural Network ANN, Naïve Bayes and Decision Tree classification methodology and HOG, LBP, PCA, SIFT, SURF, Haar–like features descriptors methodology. In the second part the research of the industrial inkjet printer sprayed information on the package information identification is performed, using the classification and descriptor combinations, which were described in the first part of the work. |
