Abstract: Traffic congestion has been one of predominant troubles that most metropolises are going through. It is assumed that identification of congestion is the first step for choosing appropriate mitigation measures. Congestion - each in belief and in reality - impacts the motion of people. Traffic congestion wastes time, strength and causes pollution. _ere are broadly two factors, which impact the congestion .congestion is ‘induced’ on the ‘micro’ stage (e.g. on the road), and ‘pushed’ on the ‘macro’ stage. _e micro degree elements are, as an example, many humans want to transport on the same time, too many motors for constrained avenue area. On the opposite facet, macro stage elements are e.g. land-use styles, vehicle possession traits, nearby monetary dynamics, and so on. Street traffic jams preserve to remain a first-rate problem in most cities round the sector, especially in growing regions ensuing in large delays, improved gasoline wastage and financial losses. Because of the poorly planned road networks, a common final results in many developing areas is the presence of small critical areas that are commonplace hot-spots for congestion; bad site traffic management around these hotspots potentially results in elongated visitors jams. in this paper, we first present a easy automatic photo processing mechanism for detecting the congestion tiers in street visitors by processing cctv digital camera photograph feeds. Our algorithm is especially designed for noisy visitors feeds with terrible picture first-class. Based on stay cctv camera feeds from multiple traffic signals in kenya and brazil, we show evidence of this congestion fall apart behaviour lasting long time-intervals across more than one places.

Keywords: traffic congestion, data collection methods, congestion measurement.