Find solution
How to consume SOAP in android?

How much resources would a Pattern recognition app for Android consume?

  • I'm planning to write an Android app that uses the device camera to take an image of a set of similar objects. The purpose of the app is to determine how many objects are present in the image (count). The app would only focus on one type of objects (Ex: people or cells). What would be the better option; computation within the device or sending the image over to a remote server? If the computation can be handled by the device, would a first generation device running on version 2.2 with 528MHz of CPU and 288MB of RAM be able to return an output within a convenient time?

  • Answer:

    Image recognition is generally CPU and RAM intensive.   Normally I wouldn't worry about speed at all until you've built your first prototype, but in this case I would do the image processing in the cloud because : It's known to be a resource hog. You may wish to use a library not in Java (like OpenCV or something.) As you make improvements to it (now it can count cars, bicycles and beer cans!) all of your users get that functionality without updating their app.

Christopher Reiss at Quora Visit the source

Was this solution helpful to you?

Other answers

There are simple algorithms that can run directly on phones (simple features and linear classifiers) and more complex machine learning algorithms that take on the order of minutes per image (usually academic prototypes that try something ambitious like recognize many objects at once). You should first worry about getting an algorithm to do something useful at all. For this first stage, you'll want to develop on a modern computer. Once you have an algorithm which recognizes the patterns you care about, determine if it is fast enough to port to a phone or launch a cloud-based micro service for processing.) If your algorithm takes 1-3 seconds per image on a desktop, you should be able to perform optimizations (lots of good vision gurus out there to help you) and make it an order of magnitude faster. You shouldn't worry about premature optimization. I've seen many students start out with Android/iOS for vision projects, and so much time is spent on just making it run vs. actually making the pattern recognition algorithm better. Good luck!

Tomasz Malisiewicz

Related Q & A:

Just Added Q & A:

Find solution

For every problem there is a solution! Proved by Solucija.

  • Got an issue and looking for advice?

  • Ask Solucija to search every corner of the Web for help.

  • Get workable solutions and helpful tips in a moment.

Just ask Solucija about an issue you face and immediately get a list of ready solutions, answers and tips from other Internet users. We always provide the most suitable and complete answer to your question at the top, along with a few good alternatives below.