There are various methods to determine the number of people in a gathering but shortfalls exist in random estimation for open air ones; ticket sales and sit numbers increase accuracy for some while aerial photography does for others.
These methods are rarely exact and remain as estimation even when accuracy is desired. People gather for different concerns showing support, increase or involvement. In more than just estimation, the exact number of people in large gatherings is important.
Video cameras and high definition cameras (of late) are mostly used to cover large gatherings. Picture quality and other features give good images but an important feature (Crowd Counting) that can be subsumed is unavailable.
Using face detection technology, a camera can detect faces in a crowd; numbering and saving them as it zoom and move from faces. Returning to a previous area, face recognition technology will recognize a detected face from a library of saved faces to avoid recounting.
This solution is what “Digital Camera for Crowd Counting” a research paper available for download presents. It describes in details how this feature can come with new cameras raising the number of solution cameras hand when in use at large gatherings; this will also improve aerial photography counts accuracy.
This solution will come to mobile phones after substantial integration in cameras; it is completely different from people counter technology in malls and sole face detection feature in digital cameras because it lets coverage and counting go simultaneously. Big players like Sony and Toshiba may want this first but Google and Apple may develop the algorithms because of their successful records in news technologies.