Infinite Coffee Cup

Simha, Sagar N and Kumar, Manoj M and Nagaraja, Vishaal and Prabhakar, T and Jamadagni, HS (2013) Infinite Coffee Cup. In: TEXAS INSTRUMENTS INDIA EDUCATORS' CONFERENCE (TIIEC 2013) , APR 04-05, 2013 , Bangalore, INDIA , pp. 84-90.

PDF IEEE_Tex_Ins_Ind_Edu_Con_84_2013.pdf - Published Version Restricted to Registered users only Download (661kB) | Request a copy

Abstract

Imagine you go to your favorite cafe house or a restaurant and the system recognizes your presence and alerts the waiter and caters to your favorite beverage you longed for the day. The main motivation to Infinite Coffee Cup is to detect the presence of a person and dynamically and non-Invasively estimate the amount of coffee in a circular cup on the table. One of the more popular devices used is Microsoft Kinect, which has cameras that capture both RGB and Depth data. Kinect is a low cost sensing device which provides two streams of images, 8 bit 3 channel RGB image and 11 bit single channel depth image. The system is fixed at the roof facing downwards and it takes a top view of the cup placed on the table. We use ``libfreenect'' 1] drivers which is an open source project, to stream data from kinect. We employ various image processing algorithms using OpenCV (Open Source Computer Vision) which is a library of programming functions for real time computer vision. The algorithms are used to manipulate the data from kinect to, - Detect the circular coffee cups in the RGB image. - Overlay and correlate the depth values of cups in the depth image to get the height and amount of coffee in the cup.

Item Type:	Conference Proceedings
Publisher:	IEEE
Additional Information:	Copy right for this article belongs to the Texas Instruments India Educators' Conference (TIIEC), Bangalore, INDIA, APR 04-05, 2013
Department/Centre:	Division of Electrical Sciences > Electronic Systems Engineering (Formerly Centre for Electronic Design & Technology)
Date Deposited:	24 Aug 2016 09:43
Last Modified:	24 Aug 2016 09:43
URI:	http://eprints.iisc.ac.in/id/eprint/54578

Actions (login required)

View Item