The deep learning technique has proven to be effective in the classification and localization of objects on the image or ground plane over time. The strength of the technique's features has enabled researchers to analyze object trajectories across multiple cameras for online multi-object tracking (MOT) systems. In the past five years, these technical features have gained a reputation in handling several real-time multiple object tracking challenges. This contributed to the increasing number of proposed deep learning methods (DLMs) and networks seen by the computer vision community. The technique efficiently handled various challenges in real-time MOT systems and improved overall tracking performance. However, it experienced difficulties in the detection and tracking of objects in overcrowded scenes and motion variations and confused appearance variations. Therefore, in this paper, we summarize and analyze the 95 contributions made in the past five years on deep learning-based online MOT methods and networks that rank highest in the public benchmark. We review their expedition, performance, advantages, and challenges under different experimental setups and tracking conditions. We also further categorize these methods and networks into four main themes: Online MOT Based Detection Quality and Associations, Real-Time MOT with High-Speed Tracking and Low Computational Costs, Modeling Target Uncertainty in Online MOT, and Deep Convolutional Neural Network (DCNN), Affinity and Data Association. Finally, we discuss the ongoing challenges and directions for future research.

Multi-object tracking in video surveillance is subjected to illumination variation, blurring, motion, and similarity variations during the identification process in real-world practice. The previously proposed applications have difficulties in learning the appearances and differentiating the objects from sundry detections. They mostly rely heavily on local features and tend to lose vital global structured features such as contour features. This contributes to their inability to accurately detect, classify or distinguish the fooling images. In this paper, we propose a paradigm aimed at eliminating these tracking difficulties by enhancing the detection quality rate through the combination of a convolutional neural network (CNN) and a histogram of oriented gradient (HOG) descriptor. We trained the algorithm with an input of 120 × 32 images size and cleaned and converted them into binary for reducing the numbers of false positives. In testing, we eliminated the background on frames size and applied morphological operations and Laplacian of Gaussian model (LOG) mixture after blobs. The images further underwent feature extraction and computation with the HOG descriptor to simplify the structural information of the objects in the captured video images. We stored the appearance features in an array and passed them into the network (CNN) for further processing. We have applied and evaluated our algorithm for real-time multiple object tracking on various city streets using EPFL multi-camera pedestrian datasets. The experimental results illustrate that our proposed technique improves the detection rate and data associations. Our algorithm outperformed the online state-of-the-art approach by recording the highest in precisions and specificity rates.

The increase in security threats and a huge demand for smart transportation applications for vehicle identification and tracking with multiple non-overlapping cameras have gained a lot of attention. Moreover, extracting meaningful and semantic vehicle information has become an adventurous task, with frameworks deployed on different domains to scan features independently. Furthermore, approach identification and tracking processes have largely relied on one or two vehicle characteristics. They have managed to achieve a high detection quality rate and accuracy using Inception ResNet and pre-trained models but have had limitations on handling moving vehicle classes and were not suitable for real-time tracking. Additionally, the complexity and diverse characteristics of vehicles made the algorithms impossible to efficiently distinguish and match vehicle tracklets across non-overlapping cameras. Therefore, to disambiguate these features, we propose to implement a Ternion stream deep convolutional neural network (TSDCNN) over non-overlapping cameras and combine all key vehicle features such as shape, license plate number, and optical character recognition (OCR). Then jointly investigate the strategic analysis of visual vehicle information to find and identify vehicles in multiple non-overlapping views of algorithms. As a result, the proposed algorithm improved the recognition quality rate and recorded a remarkable overall performance, outperforming the current online state-of-the-art paradigm by 0.28% and 1.70%, respectively, on vehicle rear view (VRV) and Veri776 datasets.