Abstract
This paper presents a low-cost multiple object tracking (MOT) technique by employing a novel appearance update model for object appearance modeling using K-means. The state-of-the-art work has attained a very high accuracy without considering the real-time aspects necessitated by currently trending embedded vision platforms. The major research on multiple object tracking is used to update the appearance model in every frame while discounting its persistent nature. The proposed appearance update model reduces the computational cost of the state-of-the-art MOT 6-fold by exploiting this facet of persistent appearance over the sequence of frames. To ensure accuracy, the proposed model is tested on different publicly available standard datasets with challenging situations for both indoor and outdoor scenarios. The experimental results illustrate that our model successfully achieves multiple object tracking while coping with long-term and complete occlusion. The proposed method achieves the same accuracy in comparison with the state-of-the-art baseline methods. Moreover, and most importantly, the proposed method is cost-effective in terms of computing and/or memory requirements in comparison to the state-of-the-art techniques. All these traits make our design very suitable for real-time and embedded video surveillance applications with low computing/memory resources.
Highlights
Multiple object tracking (MOT) is the process of estimating the state of moving objects in a sequence of video frames captured using a camera. It is a method of segmenting and observing an object’s spatiotemporal modifications, i.e. to keep track of its presence, orientation, shape, motion, position, size, and occlusion, and to extract context information, which is useful for higher-level applications
We present a novel object appearance update model for multiple object tracking and the proposed model can be applied to any MOT technique for significant reduction in computational cost
Our model has achieved the same accuracy as the baseline appearance-based MOT algorithms of Gaussian mixture model (GMM), histogram, and K-means for standard datasets
Summary
Multiple object tracking (MOT) is the process of estimating the state of moving objects in a sequence of video frames captured using a camera. Tao et al [23] gave a proposal based on motion, appearance, and shape models to accomplish highly precise MOT by employing an EM algorithm This method has fairly high computational complexity as it requires estimation of different model components and parameters. In our previous work [26], we designed a novel K-means-based appearance model for MOT to address both the issues of memory requirements and computational cost. We present a novel object appearance update model for multiple object tracking and the proposed model can be applied to any MOT technique (e.g., histogram-based MOT, GM- based MOT, or K-means-based MOT) for significant reduction in computational cost.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have