CUDA Accelerated X-ray Image Processing for Bone Detection in the Valka X-ray Guided Cutting Machine

The Valka X-ray Guided Cutting Machine is a fish fillet cutting machine which takes an X-Ray image of every fillet fed to the machine via conveyor. The x-Ray is then joined with a 3D image which gives a three dimensional image of the fillet and its bones. The cutting is made with high pressure water jets. The water jets are controlled by robots which cut out the bones and portion the fillet with high precision while the fillet is rolling on the conveyor. The result is precision cut fillets for maximum revenues. This project involves rewriting the existing bone detection for the Valka X-Ray Guided Cutting Machine using CUDA to further accelerate the bone detection allowing more time for other calculations such as cutting analysis. The new CUDA based bone detection is then compared to its CPU based predecessor for benchmarking. It also involves researching if OpenCV is a viable option for future development of image processing and recognition for Valka machines.

Authors dc:creator

• Emil Atli Ellegaard 1988-
• Sveinn Elmar Magnússon 1978-

Contributors dc:contributor

• Háskólinn í Reykjavík

• Tölvunarfræði
• Tölvufræði
• Sjálfvirkni
• Röntgentækni
• Computer science
• Automation
• Three-dimensional imaging
• X-rays

Language dc:language.iso

en

Handle dc:identifier.uri

http://hdl.handle.net/1946/22125

OAI identifier oai:identifier

oai:skemman.is:1946/22125