I am currently doing thesis on license plate recognition of Myanmar Number Plates. But I have encountered some error with it. I have to read two lines of characters from the plate but in my result, only one line appears. I used bounding box analysis for character segmentation and all characters on the number plate have been bounded. However, when extracting the bounded characters one by one , the characters on the upperline do not appear. So, how can I do for extraction of all characters. The screenshots and m file including bounding box analysis are attached below. One more thing to ask is that how can I improve the result of OCR using template matching since I have got some incorrect characters in recognition.Thank in advance.