Multimodal (RGB and thermal) applications are swiftly gaining importance in the computer vision community with advancements in self-driving cars, robotics, Internet of Things, and surveillance applications. Both the modalities have complementary performance depending on illumination constraints. Hence, a judicious combination of both modalities will result in robust RGBT systems capable of all-day all-weather applications. Several studies have been proposed in the literature for integrating the multimodal sensor data for object tracking applications. Most of the proposed networks try to delineate the information into modality-specific and modality shared features and attempt to exploit the modality shared features in enhancing the modality specific information. In this work, we propose a novel perspective to this problem using a Siamese inspired network architecture. We design a custom Siamese cross-domain tracker architecture and fuse it with a mean shift tracker to drastically reduce the computational complexity. We also propose a constant false alarm rate inspired coasting architecture to cater for real-time track loss scenarios. The proposed method presents a complete and robust solution for object tracking across domains with seamless track handover for all-day all-weather operation. The algorithm is successfully implemented on a Jetson-Nano, the smallest graphics processing unit (GPU) board offered by NVIDIA Corporation.

Multi-modal systems extract information about the environment using specialized sensors that are optimized based on the wavelength of the phenomenology and material interactions. To maximize the entropy, complementary systems operating in regions of non-overlapping wavelengths are optimal. VIS-IR (Visible-Infrared) systems have been at the forefront of multi-modal fusion research and are used extensively to represent information in all-day all-weather applications. Prior to image fusion, the image pairs have to be properly registered and mapped to a common resolution palette. However, due to differences in the device physics of image capture, information from VIS-IR sensors cannot be directly correlated, which is a major bottleneck for this area of research. In the absence of camera metadata, image registration is performed manually, which is not practical for large datasets. Most of the work published in this area assumes calibrated sensors and the availability of camera metadata providing registered image pairs, which limits the generalization capability of these systems. In this work, we propose a novel end-to-end pipeline termed DeepFusion for image registration and fusion. Firstly, we design a recursive crop and scale wavelet spectral decomposition (WSD) algorithm for automatically extracting the patch of visible data representing the thermal information. After data extraction, both the images are registered to a common resolution palette and forwarded to the DNN for image fusion. The fusion performance of the proposed pipeline is compared and quantified with state-of-the-art classical and DNN architectures for open-source and custom datasets demonstrating the efficacy of the pipeline. Furthermore, we also propose a novel keypoint-based metric for quantifying the quality of fused output.