- Use of OpenCV to detect Russian car plates in images, and TesseractOCR (optical character recognition engine) to extract number and text from the detected car plate
When we talk about AI, computer vision is definitely one of the top applications that comes to people's minds. Hype aside, I have always been fascinated by how computer vision works, especially since it pertains to the human sight, arguably the most important organ of human sense. I have already come across plenty of computer vision projects involving the human face and/or body (e.g. eye detection, frontal face detection etc.). As such, I decided to pursue a project on car license plates instead, which appears to be slightly less explored. Another important reason why I worked on car plates is that after detection of the license plate, it allows me to transition very nicely into the application of optical character recognition (OCR) to extract and display the string of text from the identified car plate.
- Link to Medium article: https://towardsdatascience.com/russian-car-plate-detection-with-opencv-and-tesseractocr-dce3d3f9ff5c
- Link to Jupyter Notebook NBViewer: https://nbviewer.jupyter.org/github/kennethleungty/Car-Plate-Detection-OpenCV-TesseractOCR/blob/main/Car%20Plate%20Detection%20with%20OpenCV%20and%20TesseractOCR.ipynb