Noise emission from individual vehicle largely contributes to city pollution and has serious health impact. The standards towards vehicle manufactures consists in pass-by testing with specific acceleration conditions which are not representative of all real driving. For the 2015/996 EU directive, the vehicle source model is inspired of the preceding pass-by standard with derived data represented the propulsion and rolling noise sources. Anyhow, those sources are underestimated due to driving behavior, aged and modified vehicle, road surface, meteorological conditions The true data collection of vehicle pass-by would be interesting. Moreover, some of the countries are reflecting on how to fight against those extremely noisy vehicle exceeding noise limit with efficient monitoring systems. This paper presents an innovative tool able to detect, identify and quantify the noise emission of individual pass-by vehicle in real life traffic. It is based on the combination of array and video processing. Compared to the state of the art and thanks to MEMs technology, the system is optimized and designed to quantify the individual noise vehicle emission regarding standard with controlled measurement and accurate processing. If the conditions are not respected to properly qualify the pass-by regarding the system limits, the data are ignored. It aims at constructing large and accurate database useful to determine average noise levels and/or acceptable noise limits per vehicle category.