Traffic congestion is a crucial issue that raises the uncertainty level of the traveling duration resulting in high stress and unsafe traffic scenarios. Effective traffic estimation and forecasting via Intelligent Transportation Systems (ITS) applications are beneficial in a variety of applications. The process of accurately and rapidly predicting the traffic condition helps the travelers to determine the traveling path and make decisions wisely. This paper develops a new deep learning (DL) based traffic density estimation and prediction model for ITS. The proposed model involves a set of two DL models, namely convolutional neural network (CNN) and long short term memory (LSTM) for traffic density estimation and prediction. These models are applied, and the results are analyzed under diverse situations. The experimental outcome indicated that the LSTM model is superior to CNN on both estimation and prediction processes.