Forecast Upcoming Traffic States by Exploiting Nearest Junctions and Big Data

Like other developing country, traffic jam is one of the common problems for Bangladesh. Every day almost a hundred thousand vehicles roaming and creating a serious traffic jam in the metropolitans. In this manner, it becomes impossible for an inhabitant to reach his/her destination on time. In this paper, we have introduced a scheme, named FTCP (Future Traffic Condition Predictor). This system has carried out to examine traffic conditions by vehicles measurement and the main theme of the study is to find out that factors which describe traffic problems by exploiting big data concept. We have categorized each key factor in different classes and each class has different traffic causes along with time durations of self-solving. When a traffic seeker is requesting for the information for a specific junction by any electronic devices, the system has acquired data information in favor of two ways. The first acquisition will be completing by collects previously recorded data and on the other, current data information of target junction via images and date-times. Here, previously recorded data were analyzed according to system define format and image samples were estimates by through FCN (Fully Convolutional Network). At last, experiments all of the major factor by vector analytical method. The experimental simulation was achieved promising accuracy in terms of using sophisticated datasets.