Qualitative Analysis of Inter-Vehicle Relationship for Scenario Parsing

Currently, the prevalent frontal collision warning systems (FCWS) are mainly based on quantitative ways. Their warning algorithms usually do prediction and assessment in the quantitative level which can not supply a universal quality under different traffic scenarios. The lack of cognition to surroundings may probably mislead the threat assessment. Besides, it is not the quantitative method but qualitative way in which people make judgments. So the scenario parsing together with qualitative methods was proposed. From this view, a qualitative analysis of inter-vehicles is presented as a step forward along the scenario parsing roadmap. The real-data experiments illustrate persuasive results.