Thesis: Applied computer vision & object detection for urban mai
Semcon Sweden AB / Datajobb / Göteborg
Observera att sista ansökningsdag har passerat.
Visa alla datajobb i Göteborg,
Mölndal,
Partille,
Kungälv,
Lerum eller i
hela Sverige Visa alla jobb hos Semcon Sweden AB i Göteborg,
Mölndal,
Härryda,
Kungsbacka,
Borås eller i
hela Sverige Semcon is now looking for 1-2 students who would like to do their thesis with us. It is important that you have a background within programming, systems and controls or mechatronics. You also have an interest for system simulation, mathematics, sensors and user experience.
Autonomous drive is a hot topic and large efforts are made on a global scale towards its full realization in various vehicle dependent sectors. A very popular area is indeed transportation, which is readily highlighted in media. The final product will have to comply with strict rules and legislation, many connected directly or indirectly to the relative high speed of the vehicle.
Another important sector is maintenance of urbanized areas such as cycle paths, and especially so during winter. This is normally performed with maintenance vehicles such as snowplows. The maintenance sector would benefit greatly from implementation of autonomous drive, which could increase operation precision, yield higher readiness, shorter deployment time, enable longer working periods and offer larger work capacity.
As the operational speed of the maintenance vehicle is low and it can be monitored by a remote operator, the introduction of autonomous vehicles is realistic even though they are expected to interact more often and closely with unprotected traffic participants (pedestrians, cyclists). To assure safe operation a number of sensors scan the vehicle surroundings to prevent and mitigate collisions and physical interaction.
The goal of this thesis is to implement computer vision using existing equipment and continuously perform object detection within the camera view. The relative positions of the detected object(s) will be published on the internal system to be used by vehicle control logics. In this case two cameras are used, front and rear facing. The usefulness of stereo vision will also be explored, documented and possibly implemented. One of the challenges will be to perform object detection as a lift and tilt arm of the vehicle is always at least partially in view and moving into different position during operations. For different work tasks, different tools are used (brush, plow, fork) which may obstruct the view differently. As a complement to computer vision, also radar and ultra-sonic technologies will be investigated, and combined with sensor fusion methodologies. Optionally further usages of the sensors can be investigated: The cameras will serve several other functions such as enabling navigation using image data when GNSS navigation fails and provide a live view for the operator interfacing with the vehicle from a remote location. This live view may include augmented reality functions such as projected path trajectory and other operation critical data.
Please apply as soon as possible.
Contact:
Team Manager, Anne Piegsa
anne.piegsa@semcon.com46 739 415 770
Varaktighet, arbetstid
Heltid Tillsvidare
Publiceringsdatum2019-11-06ErsättningLön enligt överenskommelse
Så ansöker duSista dag att ansöka är 2019-12-05
Klicka på denna länk för att göra din ansökanFöretagSemcon Sweden AB
Omfattning Detta är ett heltidsjobb.
Arbetsgivare Semcon Sweden AB (org.nr 556555-8193)
Arbetsplats Semcon Sweden
Jobbnummer 4943221
Observera att sista ansökningsdag har passerat.