My project is detect human fall in elderly people. So I'm using latest OpenCV in python library and for human pose estimation i'm using Humaneva dataset. I am new to opencv. So, anyone can tell me what will be the procedure to detect human Fall.
definitely not an easy, or even a beginner's project. don't expect to find anything ready-made, save using opencv's python api.
then, you don't tell us anything about the constraints here. video ? single images ? outdoor ?
if you did any reasearch on this before, why doesn't it show up in your question ?
this looks quite good , but again, advanced !
@berak I want to detect human fall in videos. Yes I've seen some papers on it but idea is not that much clear in it. So I decided to firstly detect human posture using humaneva dataset. Then we will go to detect fall position with the help classifier. But I'm currently stuck in humaneva dataset. It shows me core dump error. I've posted the problem in question how to run humaneva dataset in opencv later in comment.
" I've tried with haar cascade dataset" -- won't work. both the cascades and the HOGDescriptors were trained with upright people.
also, already for upright persons, shadows are a problem. for a person on the ground, this will be a terrible problem. any chance, you can use more sophisticated hardware, like a depth cam ?
Asked: 2018-01-18 03:59:37 -0600
Seen: 1,171 times
Last updated: Jan 18 '18
