Stilling:
Summerjob - INDISAL project
Application due: 18.03.2019
Computer Vision and Machine Learning meets Aquaculture
In the near future Computer Vision and Machine Learning based decision support systems will help Norwegian fish farmers to observe the state of farmed fish and will enable the farmer to make optimal decisions regarding feeding, fish welfare and other operational related treatments.
Due to an increasing availability of high quality image and video-data within aquaculture it becomes important to exploit and understand a huge amount of recorded data.
For the training and evaluation of many supervised and unsupervised machine learning approaches it is often necessary to label image and video-data. As this is a labor intense task we wish to make this process more efficient and convenient by the creation of OpenCV based helper tools.
The SINTEF-Summer-student will help to design, create or extend such C++/OpenCV based programming tools. The aim is to make them generic enough such that they easily can be extended or configured for different aquaculture applications using a configuration file.
Keywords:
Computer Vision, Deep-Learning, C++, OpenCV
The expected work tasks are:
- Design/Create an OpenCV/C++ based tool that allows to label video and image data in a convenient and configurable way
- Allow for an easy integration of tracking and recognition modules
- Integrate a deep-learning recognition/segmentation module
- A short user friendly description (including example code) should be created
- Evaluate and improve the tool based on example cases
Requirements:
- Interest to apply Computer Vision and Machine Learning techniques in aquaculture applications
- Good C++ programming/documentation skills
- Knowledge or experience of OpenCV
- Operation system(s): Linux (and Windows)
- Knowledge/experience with cmake, git, doxygen is a plus
Web: www.sintef.no
Kontakt: Christian Schellewald (Forsker)
Telefon: +47 92280980
E-post: Christian.Schellewald@sintef.no