We need to count the number of people in the crowd, and we used opencv_traincascade , but we find a lot of things.
opencv_traincascade
What could be another approach that we can use?
I was looking for something like that. I found this HeadCounter project ( https://github.com/yquemener/HeadCounter ) and some documents: "Counting in extremely dense crowd pictures" ( http://crcv.ucf.edu/projects/crowdCounting/index.php ) and "Crowd Size Estimation" ( http://course1.winona.edu/cmalone/promotion/UndergraduateResearch/Cariveau%20Report.pdf )
There are several ways to do this. I did a similar project a few years ago using Aforge.net.
I suggest you use some existing image library such as emgu-cv or aforge for this task.
I think the http://www.codeproject.com/Articles/239849/Multiple-face-detection-and-recognition-in-real-ti article can help you achieve this functionality with a sample project.