I need a fast and accurate method that allows, from the capture of a standart webcam (640x480), detect a face in the scene (frontal but may have an angle) and determine if the eyes are open or closed, or better yet, how open they are.
* You can rely on a previous calibration from the user's photo (eyes open and closed), but basically it should work without such calibration, or rely on it more for fine tuning
* The distance, and therefore, the size of the face and eyes may vary (the maximum distance from the camera will be 2 meters)
* The system must adjust the brightness of the image automatically to correctly detect the eyes in different light conditions.
* the class must provide face and eye position and whether they are open or closed
+ compiled class is required and properly documented source code.
+ final program must run on a mobile device (android or windows mobile), so the software must be compatible.
+ a second part of this project may require detection of gestures
6 freelancers are bidding on average $650 for this job
I am an expert working with vision systems using open cv. As a beginner in this website I will get this project done for 400$. As you give the details i can tell you how accurately this can be done.
We are Shen Zhen based website company, we have work a similiar project before by OpenCV. We are looking forward to cooperate with yours. Pls check the PMB for face dtection demo.
Hi, I am new to freelancer, so not much details you can get about me here. But I can share more details about me and the work done if you are interested.