Click here to Skip to main content
15,793,609 members
Please Sign up or sign in to vote.
0.00/5 (No votes)
I want to detect an object which would be fixed and take a picture of that object using Android Camera. I searched google for it and found openCV which seems to be optimal result but the problem is there is no exact help for how to go about it. I have already tested the samples that are available in the openCV package. Also the samples like face detection are not working. If anyone has tried any such similar thing please help. I also found BoofCV but hw to go about it is a quiz. OpenCV looks to be developed using c/C++ for which I am new. Any input in this regard will be helpful.

1 solution

Blessings - image analysis is not something for which there is a clear solution nor something that can be patched up quickly.

OpenCV has a sample project with autolearning for face recognition though, it needs to be taught and I can't say if it's good or bad - I work in image analysis but from a different perspective.

You should find some good books on Image Processing like "Digital Image Processing: PIKS Inside, Third Edition. William K. Pratt" (ISBN: 0-471-37407-5) or "The image processing handbook, John C. Russ" (ISBN: 0‑8493‑7254‑2) to get a hold of the main problematics with this field.

Also, delve deep into OpenCV documentation, to have a preliminary knowledge of the problem it took me almost a week.

Face recognition in particular has deep roots in Artificial Intelligence because algorithmic, deterministic approaches tend to be unreliable due to the sheer number of small variations in the image processed that can completely mislead the algorithms.

EDIT: sorry, I was talking about face detection and you were asking for object detection.

In this case it's even worst than face detection because you have to know what you're looking for. The best way to start implementing object recognition is to turn the image in graylevel and woprk your way out with Sobel or Laplace operators to get the edges of the image. From then on it is a matter of logic to discriminate objects with respect to background, similar objects, junctions of objects and so on.

Hope to be of help,
Share this answer
Member 11233687 14-Nov-14 7:42am    
the object is in fact a cooler used for keeping cold drink bottles.
den2k88 14-Nov-14 7:47am    
So you know it's shape. I'd say that either you try with the automatic learning project of OpenCV (I don't remember its name and I don't have OpenCV doc at hand nor free internet access, sorry for that) and a large number of samples - that would be the easy way since it requires little coding and headache or you start designing the logic to understand, based on the edges/contrasts/shape of the image if you are looking at a cooler o not.
That requires experience and a not trivial work to be done, but there are no other ways. Image processing is largely a non deterministic field.

EDIT: I found an answer on stackoverflow HERE

There are a little more accurate pointers than the ones I gave you before.
Member 11233687 14-Nov-14 8:25am    
thanks for the input I am absolutely new to image processing and have only been working on Android apps but this is a very new task to me as well. i will try and work on these and will let u know.

This content, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)

CodeProject, 20 Bay Street, 11th Floor Toronto, Ontario, Canada M5J 2N8 +1 (416) 849-8900