Vehicle density estimation plays an important role in efficiently reducing traffic since it assists in making time-space adjustments on real-time traffic flows, thereby preventing congestion of roads. Within the last 5 years, average statistics show a rise of 11.3% in the traffic congestion index, corresponding to an increase of 7% in accidents in our target city. This approach goes beyond distinguishing between two-wheelers and four-wheelers; it can count pedestrians, hence allowing the estimation of their density. By utilizing the advanced capabilities of vehicle counting techniques, it is simple to find out vehicle density fairly and enhance the traffic management strategy. The Non-maximum Suppression algorithm (NMS) used here has an advancement where accuracy for counting vehicles is improved by eliminating duplicate detections and leaving the most confident ones. In contrast, Hidden Markov Model (HMM) estimates vehicle density accurately by relating traffic patterns and vehicle counts. This fully integrated traffic management and urban planning approach adds more to the smart city agenda through data-driven insight for optimum infrastructure development and strategic urban planning. The performance metrics used in appraising the model are Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE), which give an 8.5% MAE and 8.77% RMSE result, respectively.