Visual Place Recognition with Low-Resolution Images
This work addresses the problem of enabling visual place recognition on low-end commercial systems with resource constraints, but it is incremental as it focuses on analyzing existing handcrafted methods rather than introducing new ones.
The paper analyzes how image resolution affects the accuracy and robustness of handcrafted visual place recognition pipelines, finding that these low-computation methods can adapt to flexible resolutions, making them suitable for resource-limited systems.
Images incorporate a wealth of information from a robot's surroundings. With the widespread availability of compact cameras, visual information has become increasingly popular for addressing the localisation problem, which is then termed as Visual Place Recognition (VPR). While many applications use high-resolution cameras and high-end systems to achieve optimal place-matching performance, low-end commercial systems face limitations due to resource constraints and relatively low-resolution and low-quality cameras. In this paper, we analyse the effects of image resolution on the accuracy and robustness of well-established handcrafted VPR pipelines. Handcrafted designs have low computational demands and can adapt to flexible image resolutions, making them a suitable approach to scale to any image source and to operate under resource limitations. This paper aims to help academic researchers and companies in the hardware and software industry co-design VPR solutions and expand the use of VPR algorithms in commercial products.