Detection of Traffic Congestion from Surveillance Videos using Machine Learning Techniques

Smart Cities applications, automated traffic control and management is the most trending research area. With the improving needs of developed towns and cities traffic congestion, now a days this the traffic congestion control and its applications has large needed facing problem in the increased population cities. Peeled eye camera photos and videos can be watched efficiently to detect traffic congestions in most of the areas in the grown populated cities. The earlier researchers had observed more on traffic signal controls through photos executed by using different algorithms of machine learning. There is existing research available on traffic signal controls through image processing and various machine learning methods. The image features are extracted and interpreted for the same. Deep learning algorithm, convolutional neural network (CNN) is proposed for effective detection of traffic congestion. There were existing works available in traffic detection using machine learning and deep learning approaches. Machine learning, Nowadays, traffic surveillance systems collect a lot of videos or images and store them for the live monitoring purposes. Deep learning techniques are used sparingly in traffic surveillance and control systems. Various images with various weather conditions are collected and are used as training dataset. Advantages of deep learning have been exploited in many applications, which use computer vision and image analysis. One of such applications is traffic monitoring, in which large amounts of video or images are processed for effective learning. The traffic surveillance can only monitor, which cannot detect the traffic on particular time.