Chenhui2018
commented
7 years ago Hello @hoangtocdo90, did you solve the problem? As far as I know, the "trained_classifierNM1.xml" was created by Boost with 8 features, you can check the paper
"Real-Time Scene Text Localization and Recognition - Luka ´s Neumann Ji ˇ ˇ r´ ı Matas"
for detail. free to download.
My problem is what is the training set for "trained_classifierNM1.xml". any information is helpful, thanks for sharing!
Hi guys! thank for come in 🗡 i'm new for opencv. i'm testing opencv.text module with tesseract. Everythings is ok but when i try to process with non english ocr images, doesn't catch any non english char, i;m thinking maybe need train some things in this xml file. Some body can tell me how to create it for non english. Thanks you ~~