Machine Learning Based Medicine Distribution System
Huiling Xia, Chunzhi Wang, Lingyu Yan, Xinhua Dong, Yichao Wang · 2019
In this thesis, an intelligent medicine recognition method for medicine vending machines has been proposed. Considering that the name of the medicine is generally the largest character in the medicine box, the medicine name recognition can be converted into the largest character recognition in the picture in this thesis. First, we used Support Vector Machine (SVM) and Connected Component to determine the text region, and found the largest connected region to locate the medicine name. Secondly, we use the method of “fragment link” for text segmentation, which divides the text into two elements: fragment and link, and combines the whole word according to the set rules. Finally, we can directly use Optical Character Recognition (OCR)software for character recognition. Experiments show that this method has a high accuracy of medicine identification.