Image segmentation: how to isolate human body figure and a complex background? - image-processing

I need do it without user intervention. The user upload a full body image and the server must change the background automatically. I think I will use the ImageMagik to change the background, but I have no idea how to remove the background. I am trying opencv+graphcut but i am not sure if is a solution. I know it's difficult but I have to find a feasible solution. Maybe you know some service to do this. You all can give any approach or suggestions that will make a real help to me. thanks in advance!

If you are working with complex background then you should try GrabCut. It is interactive so you need additional information from the user.
Check also "1.1 Previous approaches to interactive matting" section.

Related

An app that just reads a bar code

I want to develop an app that just reads a bar code. I don't want it to go off to a website or shop or anything. Just read it! It must run on iPad Air. I have seen zbar and a few other suggestions but I want to be sure which one I should use/how to start. Can someone who has done this type of thing help please?
Start with the sample code here that describes iOS native code detection and usage: http://www.infragistics.com/community/blogs/torrey-betts/archive/2013/10/10/scanning-barcodes-with-ios-7-objective-c.aspx
When you've progressed to needing something this doesn't accomplish, you will likely have the vocabulary to better describe what you need, what you've tried, and how it has failed for you.

Use both GPS and Scan function in same channel with Junaio AR

As the title says, im simply wondering if it is possible to use the GPS tracking and POI part of junaio, and att the same time use the scan functionality to scan and recognize images. Im working with a group at a project which demands that we use both functionalities, and we are at the moment stuck on trying to send 2 XML documents, causing the server to return nothing at all. I simply want to know if it is possible to use both functionalities in the same channel, and I would greatly appriciate if someone would point me in a direction which could help me solve our problems, since I've been able to find absolutley nothing on my own. Thanks beforehand!
Scan + GPS/compass is not possible at the moment.
However, it's possible to use GPS/compass tracking and continuous visual search at the same time. This might be the closest thing to your requirements.
You might find more information on http://helpdesk.metaio.com

asynchronous request of images by url for scrollview ios

I wanted to know how it is possible to make asynchronous request of images, and display them in a scroll view in a way that I can scroll the images without making requests for each images. I read some threads that gave me some ideas, so I think that I have to stock the urls in an array, and then I don't know what to do. If someone can explain to me ? or have a concrete idea on how to do this ?
NSScreencast has a video that does something similar. It is also a good video to learn about blocks too.
http://nsscreencast.com/episodes/10-fun-with-blocks
in order to load images asynchronously there are many options. The best approach is to create a different thread and fetch the url.
Sometimes its better to used tested and trusted apis available. I will suggest you to use SDWebImage
This will not only help you to load images asynchronously but also provides the caching mechanism. Which you will later or sooner user in you app development.
Its easy to use. Happy Coding :)

Responsive web-app design related to user behaviour

we built a webapp. We do not know who will actually use the app. Some will use it on the computer some on the iPad. Moreover one user will log in and will spent some time to accomplish all given tasks. On the iPad it is obvious that expierenced users will expect zoom and wipe effects, unexpierenced users will prefer desktop-similiar behaviour. I thought I will try to implement an algorithm which detects how expierenced a user is and the user will be categorized to the related profile.
Where can I read about this topic? Thanks in advance for some links to relevant articles.
Ok, I did some research and 'User modeling' seems like a good start.

What is the purpose of 'bouncing' in iOS?

In beta testing an iOS application I wrote, I received some feedback to let more elements 'bounce', even if there isn't more content to see, or load in.
While I don't think its necessarily a bad suggestion it got me thinking - what is the purpose of allowing this behavior on UI elements? Is this stated in the HIG somewhere that I missed?
If the list bounces you know that it is at the end instead of just frozen or something. It feels good to get feedback whenever you touch something even if nothing is happening.
While this question is better suited for http://programmers.stackexchange.com, I shall answer it the best I can here:
It gives the user feedback that something is happening
It helps to prevent the user from accidentally doing something they may not want to do (cancel editing, delete important data) etc.

Resources