Posts Tagged ‘Image processing’

Part of the innovation in replacing the default gallery application in Android is having meta-data used of generating tags that than are assign to user defined categories, sort of an automated organizer of sorts. Obviously, I can use file names as the first version of this feature.

But, with all the buzz in getting higher GHZ devices into th market not just the low end at 800 MHZ at the low end but the ones at both 1 GHZ and 1.5 GHZ cpu speeds it would seem that innovation of tagging should use image processing. For people objects that would be called face recognition, for non people object its called object recognition.

You the user choose an image on the SDcard and Gallerie calibrates the tagging of that object in the picture so your photo album has all the pictures on the Sdcard of that object in one album. with the new device cpu speeds it should be fast enough as I would be using an Android specific processing.org image library. I could even make this the premium feature and the picture tagging by file name the freemium feature.

Adobe you bore me

Posted: December 6, 2009 in Android, Java, Mobile
Tags: ,

I would imagine many of you have tried the mobile application from Adobe for Android. Its okay I guess but I want to go in a different direction. Some research groups have created new languages that cover the image processing area. One of those groups has started an Andorid branch, namely processing.org, see the android source branch/trunk.

A less complex approach seems to be JJIL which just focuses on image processing itself. The created image processing pipe-lines than can be used both in processing still images and video. The actual math-programming  functions used seems to be based on OpenCV.

What I am attempting to state is there is big difference between put something up and engaging the mobile Android user. Being able to crop and color correct an image is not exactly engaging that mobile Android user. Thus the general idea is to start with an Android Image Gallery application and use JJIL to do some image processing via image processing pipelines and combine that with some new UI ideas I have to see if I can raise the engage the mobile Android user bar higher than Adobe did.

Reblog this post [with Zemanta]