APPN ANU Node staff, Ming-Dao Chia, Tao Hu and Saswat Panda, recently celebrated the success of their two TechLauncher student teams who successfully presented their projects at the ANU Computing Showcase.
TechLauncher CCS is an initiative by the ANU College of Engineering, Computing & Cybernetics (CECC) which enables students to develop research and professional skills in a real-world environment while bringing great ideas to life.
Back in February, APPN pitched two projects and attracted 14 students to work in two teams on developing AI-enabled infrared imaging systems (led by Saswat Panda) and building tools to improve the spatial resolution of hyperspectral images (led by Tao Hu).
In the infrared imaging project, students developed an automated plant image capture from RGBD and thermal cameras. They then calibrated the two cameras with each other and generated plant masks to detect individual leaves in thermal images. This work was achieved by combining data creation and labelling with Artificial Intelligence. Watch this video to learn more.
The Hyperspectral Image Super-Resolution (HSI-SR) team aimed to improve the spatial resolution of hyperspectral images to better capture fine details in plant images.
Hyperspectral images typically exhibit high spectral resolution but suffer from comparatively poor spatial resolution. The team first conducted extensive literature research and identified the most workable state-of-the-art deep learning method. Once implemented, they doubled the spatial line resolution of the scanner from 500 to 1000 pixels. The students then used both quantitative and qualitative image quality evaluation metrics to validate their super resolution technique. Watch the showcase video for more information.
This work was supported by Dr Richard Poiré and Prof Danielle Way who helped the CECC students understand the impact of technology being developed in plant study. Thanks also to the ANU AgriFood Innovation Institute and ANU CECC TechLauncher for their support of these two projects
