As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.
Closed 9 years ago.
I'm trying to implement software which automatically detects nude images. Hoping to do this through openCV. What do you think of the possibility and the best algorithm that can be used? Any examples would be highly appreciated.
Depending on how accurate it needs to be you could just define a reasonable colour space that covers flesh tones, and then have some % threshold whereby if the flesh tone pixel count exceeds this % threshold then it may well be a nude image.
You may want to keep an eye on this site. This is what chatroulette uses. It targets, ahem selective nudity, but it may be related to what you're after.
There's a downloadable demo (so far, without source).
Have you considered a neural-network approach? What you're asking for sounds like a machine-learning application.
Related
As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.
Closed 9 years ago.
I wanted to try out machine learning algorithms on iOS - to start with say, linear regression. I was advised to use octave for the PC which would make stuff like matrix operations easier. I was wondering if I could find a library for the iOS to perform those operations.
Any suggestions welcome.
Thanks.
Edit:
To not sound vague, consider solving a linear regression problem which would involve computing transpose of a matrix, inverse of a matrix. I would be happy if I have a library with API's for it. openCV does have it but using it would be completely off context then.
I'm not sure about your exact requirement on API's. But it guess it is possible to work with OpenCV on IOS and do variety of processing.
For more information please check this link.
http://docs.opencv.org/doc/tutorials/ios/image_manipulation/image_manipulation.html
Once you have image in Mat format, i think you can carry out whole lot of operation. Also the one which you have listed.
As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.
Closed 10 years ago.
The US census for 1930 is available online as a series of large pdfs. Sadly, as far as I can see, there is no free service that has also run OCR on it. There are, on the other hand, a lot of online examples of successful OCR done by "enthusiastic amateurs". So, let's start at the beginning.
For example, a single page can be seen at http://i47.tinypic.com/2i7tt8k.png, and here is another one:
What would be a good way to extract the different words from the image so that one can start to try to train a system to recognise them? Ultimately I am seeing this as a tricky machine learning problem.
You can download the entire image files from http://archive.org/details/newyorkcensus00reel1475 and associated links.
As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.
Closed 10 years ago.
I know that AI is vast field and there are many algorithms which exist there.
I have tried learning from wikipedia but thats too much math stuff which i didn't understand.
Is there any place from where i can find the 7-8 lines of brief introduction on each algorithm and which can tell me that for which type of problems we can use which algorithm.
E,g something like IF we need to find some shortest route t0o some destination then we can use shortest path algorithm. Currently there are so much variery of fields like Bayesian methods , meakov chain models , optimization algorithms which i am not able to get what type of proeblems can be solved by them.
I jsut want to know what athey are capable of doi ng rather than hard core math stuff which always goes on the top of my head
aaai.org has some good introductory articles.
http://aitopics.net/AIOverview
machine learning 101: https://stackoverflow.com/questions/598726/overwhelmed-by-machine-learning-is-there-an-ml101-book/598772#598772
As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.
Closed 10 years ago.
I am looking for a book where some monocular/visual SLAM is described and implemented.
Can you list and recommend some?
I'd like to use OpenCV but that is not a requirement.
I don't know of a book with a description of such an algorithm, but there's a complete open source implementation (in C++) of a vslam system available as part of the Robot Operating System. It uses SURF descriptors and vocabulary trees for place recognition, and bundle adjustment for SLAM. It does use OpenCV heavily as it's made by the same people. See the website here. I can't say for sure as they don't mention and I haven't looked in great detail, but their implementation seems to be based on, or at least is similar to, this paper.
Edit: The paper linked above was actually written by the people who implemented the vslam system given above, it appears. So it is definitely a good resource for understanding it.
I don't know about a book, but maybe PTAM can be useful. The ISMAR 2007 paper by Klein has a description about the system and the source code is available so you can check the details.
Of course, PTAM is just a (good) method in the SLAM field.
As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, visit the help center for guidance.
Closed 10 years ago.
I am learning digital image processing on my own and would like recomendations on good reference books. If you know of books to definately stay away from that would be useful as well.
Thanks
Digital Image Processing (!) by Gonzalez & Woods is pretty much the classic.
I used Gonzalez and Woods in my coursework, and it covers image restoration, which you say is your main interest. ''Fundamentals of Digital Image Processing'' by Jain takes a more mathematical/statistical approach and covers some different topics, but I haven't gone through the whole thing. I'd guess it'd be useful if statistics are your thing.
Probably easier would be to use Photoshop, the GIMP, or similar instead of reinventing the wheel. They can be used in batch mode in some circumstances, and you can write plugins if there's something special you want to do.
Al Bovik - The Essential Guide to Image Processing
William Pratt - Digital Image Processing
Bernd Jahne - Digital Image Processing
All of them are great, at least as good as "Digital Image Processing".