Tanzania SmartGate

In this project, autonomatic gate pass system which is called smartgate is developed with AI model. The system is consist of Object Detection Model, OCR and face recognition. For Object Detection, It would detect license plate and Container number area, and OCR would detect and recognize the results from detection. And the face recognition system is consist of face detection and face recognition. Face detection is to detect and crop the face. With this result, face recognition would compare the result to one of the list registered in the database. After comparing and verifying the identity, vehilce number, and container number, the gate would automaticaly open and the record will be stored in the database.

To accomplish this task, the object detection model is used to detect the license plate and container number region to make OCR model can read the text easily with high accuracy. For this, we test and decide to employ YOLOv11 model since it has higher speed and accuracy. Likewise, we test many different model and its combination related to the OCR and decide to use dbnet and svtr model.

To build a face recognition system, face detection and recognition model is needed. To select the decent model, we considered many factors such as accuracy, speed, computational cost, and so on. Finally, we choose to use YOLOv11 which has a better performance than others. And for face recognition, we compared many kinds of feature extraction algorithm, and head algorithm to output the proper result. At the end of many trial, we select ResNet101 for feature extraction and sub-center Arcface for head algorithm.