1. *The Android robot is reproduced and modified from work created and shared by
Google and used according to terms described in the Creative Commons 3.0
Attribution License.
*
Portable Colour Detector
Sridhar Mamella
BEng (Hons) Computers and Electronics | Academic Year - 2013 | 2014
Introduction
A method for detecting colours using a portable device with
calibration techniques such as Auto White Balance (AWB)
was devised.
Results
Digital Image Processing
• Gray Balance
• Neutral Balance
• White Balance
• K-means algorithm
Android SDK
• Eclipse + ADT plugin
• Android SDK Tools
• Android system image emulator
App
• Interfaced Camera
• RGB Values + more
• Displays colour names
• Combine versions
Initialising representatives (means)
Final App Layout
Initialise means:
• By picking k samples at random
Iterate:
• Assign each point to nearest mean
• Move “mean” to center of its cluster.
Conclusion
• A method for detecting colours using a portable device
with calibration techniques was achieved.
• Camera interfacing and digital signal processing
techniques were applied.
• Furthermore, k-means clustering was computed and
incorporated into a mobile device.
Features
• Colour detection
• Image calibration
• Identification of RGB values
• Colour storage
Android's Camera HAL overview
Android's camera Hardware Abstraction Layer (HAL)
connects the higher level camera framework APIs in android
hardware to a underlying camera driver and hardware. The
following figure describes the components involved.
Future Work
• Test more algorithms on a mobile device.
• Update colour library.
Architecture Summary