deletions | additions       diff --git a/section_Introduction_The_use_of__.tex b/section_Introduction_The_use_of__.tex  index 72376ac..457f522 100644  --- a/section_Introduction_The_use_of__.tex  +++ b/section_Introduction_The_use_of__.tex 
...
\section{Introduction}  The use of video data for automatic traffic data collection and analysis has been on an upward trend as more powerful computational tools and detection and tracking technology become available. Not only have video sensors been able to emulate inductive loops for three decades to collect basic traffic  variables such as counts and speed \cite{michalopoulos91autoscope}, but they can also provide more and more accurately higher-level information regarding road user behavior and interactions. Examples include pedestrian gait parameters , crowd dynamics and surrogate safety analysis applied to motorized and non-motorized road users ...  Among the high-level transportation applications, surrogate safety analysis has attracted particular attention this last decade and has benefited greatly from improvements in the computer vision algorithms necessary to interpret video data. Examples include vulnerable road users 
...
While flow rates and speed counts can be assessed through various means, computer vision is one of the most efficient methods of measuring other roadway interactions. Many studies are seeking to analyze microscopic movements such as \cite{St_Aubin_2013} in roundabouts and \cite{Hill_2015} in relation to driver behavior. At this scale, accurately representing the exact characteristics of the moving vehicle is crucial to provide accurate analyses of surrogate safety indicators such as gap time, time-to-collision(TTC) and user pairs. Additionally, the performance of computer vision tools are not consistent depending on several factors such as the resolution and type of camera \cite{Wan_2014} and the weather conditions \cite{Fu_2015}.  The objective of this paper is to improve the performance of feature-based tracking in merging zones of roundabouts through parameterization using a genetic algorithm comparing tracked trajectories to sets of manually annotated ground truths. There is an emphasis on cross-validating sets of parameters on separate datasets with similar properties, as well as exploring the performance on datasets with different properties. This paper will provide a brief overview of the current state of computer vision and calibration in traffic applications, then detail the methodology including the ground truth inventory, measures of performance and calibration procedure, followed by a presentation and discussion of the results and summarized in the conclusion with recommendations for future research.  big data  lack of public data, benchmarks,