How to identify which things are available in image with pymongo?

asked 2016-04-07 04:49:34 -0600

I am learning opencv from 3 days, and i want to find how many things are available in image using python coding. e.g. i have a natural photo, in this photo 3 types of different trees,some birds are flying in the sky, a dog, one bicycle and many more. now how can i identify with the help of opencv how many thing are available in this picture.

edit retag flag offensive close merge delete

Comments

2
berak gravatar imageberak ( 2016-04-07 05:03:54 -0600 )edit

Thanx @break but i want to do programming in python. Do you have any idea about algorithms?

Ghanshyam Dudhatra gravatar imageGhanshyam Dudhatra ( 2016-04-07 06:05:37 -0600 )edit
2

your language (or your favourite database) does not matter here. it's just not a "solved problem".

(all i'm trying to say is, that this is a really tough one, and may be far beyond your current skills.)

see, what you're up against: http://host.robots.ox.ac.uk:8080/pasc...

berak gravatar imageberak ( 2016-04-07 06:12:38 -0600 )edit

Good luck with that, tons of PhD students at major universities are trying to tackle this sort of problems of multiclass object detection and classification. Are you sure you have enough background knowledge to even start this project? Be ready to implement state-of-the-art algorithms for sure ...

StevenPuttemans gravatar imageStevenPuttemans ( 2016-04-08 03:45:23 -0600 )edit

currently i have 7000 negative images and 12000 positive images. i can detect object but could you please tell me which algorithm is use to detect object 100% perfectly ? i am detect object using haar/LBP cascade classifier.

Ghanshyam Dudhatra gravatar imageGhanshyam Dudhatra ( 2016-07-11 06:05:34 -0600 )edit