Hello, sir
I have worked on developing many engines using OpenCV and DNN on Embedded Linux device for image processing and audio processing for 7 years.
Project Names: license plate recognition engine, lane detection engine, face recognition engine, people counting engine, driver fatigue detection engine
Programming languages & frameworks & libraries: C++, Python, Java, Qt, OpenCV, Dlib, Yolo3, Tensorflow, Caffe, Facenet, mtcnn …
Platforms: Windows and Ubuntu desktop, Embedded Linux (JetsonTX2, Raspberry Pi, Freescale, Hi3519, DSP), Android, iOS
One year ago, I developed an engine for license plate recognition using pure CUDA C++ Yolo engine and OCR engine without any pre-built CUDA library on Windows and Jetson TX2.
I developed a people counting engine on Raspberry Pi and car-driver fatigue detection engine on Freescale(iMX 6 Quard) and Hi3519.
I have many experiences on developing the customized face recognition system (on local and Cloud server) using the machine learning and deep learning methods such as Eigenface, Fisherface and LBPH as well as facenet, arcface and lightened CNN.
Ago 20 days, I developed an app to detect, train and recognize the faces of multiple persons from live RTSP video stream.
I also developed a people counting engine at bus and store.
This engine uses head detection & tracking and face recognition.
If u want to see my additional demos, I'll send its URL.
I hope my development experience would be helpful to you.
Thanks for your attention.