Project Details
[Return to Previous Page]Autonomous surveillance of restricted area using drones - GLOBAL WITH CHALMERS UNIVERSITY
Company: AstaZero AB
Major(s):
Primary: CMPSC
Secondary: CMPEN
Optional: ME
Non-Disclosure Agreement: NO
Intellectual Property: YES
AstaZero is a proving ground located deep in the forests outside Gothenburg, Sweden built for testing of advanced automated driving, driver-assistance and active safety systems in road vehicles. AstaZero is a non-profit organization owned indirectly by the Swedish government and performs independent testing, verification and certification for its customers. An example of an active safety system could be the autonomous emergency brake system commonly found in modern vehicles. Part of track security is to ensure that the tests can be carried out in a safe manner. Unauthorized people and animals must not be in the area. The area is quite large, and an 8-10 km long fence is needed to enclose it. Occasionally the fence is patrolled by security but due to the size of the area this is not done very often. Goal In order to reduce risks, AstaZero would like to be able to use an automated drone to monitor the condition of the fences around the area. This involves full automation of the entire solution i.e. mission planning, flight, video transfer, video analysis and possibly charging. At the project’s end a design showcase will be made on the AstaZero proving ground. It will demonstrate a situation where the drone patrols a section of the perimeter, and the fence has been artificially broken. The task In this global project students from Pennsylvania State University will collaborate with students from Chalmers University of Technology in achieving the goal. Your task will be to automate video handling and analysis for fence integrity checking. The work will include the following steps: 1. Perform a cost-benefit analysis on automated fence integrity monitoring. 2. Make a technical plan for how the different pieces of the system should interact. 3. Automate video transfer from the drone. 4. Find a suitable image analysis tool. 5. Implement the solution detecting fence breaks in the video data.