This content is not included in your SAE MOBILUS subscription, or you are not logged in.
Vision based solution for auto- maneuvering of vehicle for emerging market
ISSN: 0148-7191, e-ISSN: 2688-3627
To be published on November 21, 2019 by SAE International in United States
Event: NuGen Summit
Vision based solution for auto- maneuvering of vehicle for emerging market: Author/Co-Author: Singh Ashwani, SDV Ram Kumar, Bose Souvik, Lalwani Chandraprakash General Motors Technical Centre India Key words: Image Processing, Gap finding, virtual/Imaginary lines, Advance Driver Assist System (ADAS), Vehicle to vehicle(V2V)/Vehicle to Infrastructure(V2I/V2X) Research & Engineering Objective: For the various levels of autonomous, the current perception algorithms involve considerable number of sensor inputs like cameras, radars and Lidars and their fusion logics. The planning route for the vehicle navigation is done through map information which is highly volatile and keep changing many at times. Existing steering assist feature during a curve is available by combining additional driver monitoring camera & 360 degree camera. The complexity is very high in the implementation and computation of these algorithm. These solutions are not cost-effective for emerging markets. Non-availability of required infrastructure in developing countries is one of the additional constrain. Feature unavailability due to road infrastructure (ex: poor or no lane markings), bad weather will lead to higher customer dissatisfaction. The objective is to develop a logic/study by using image processing, to derive an efficient perception of the environment for better availability of feature. Methodology: The solution presented in this paper is more towards camera based. Using camera efficiencies and advanced deep learning techniques, the perception can be derived and computed faster. The objective is to develop a low-cost predictive path with only camera for emerging market. Image processing to determine the predictive space to navigate/steering through it Result: Paper will be presented describing the method to determine the predictive path, that is obstacle free space and use the predictive space to navigate/steer. Limitations of this study: The proposed solution may need inputs from V2V /V2I/V2X functionality. Implementation of V2V functionality is out-of-scope in this paper What does the paper offer that is new in the field including in comparison to other work by the authors: Vehicle maneuverability with poor road infrastructure by using only camera Conclusions: Low Cost solution for ADAS in emerging market.