Abstract
The School of Geodesy and Geomatics at Wuhan University has developed a GNSS/SINS post-processing service system named POSMind. With the growing demand for mobile mapping in scientific and engineering applications, such as earth observation and high-precision mapping, there is a crucial need for efficient and accurate direct georeference based on GNSS/SINS integration. To accommodate diverse applications, an analysis of existing service forms was conducted, culminating the development of the hierarchical post-processing service system. This system consists of three service forms: module for interface calls, software for fine processing, and web for efficient cluster processing. POSMind has assimilated existing excellent methodologies and constructed a high-precision GNSS/SINS integration algorithm framework through theoretical derivations and experimental tests. Refinements have been introduced in several facets, including pre-processing, quality control, ambiguity resolution, and smoothing schemes. To assess the performance of POSMind, a series of experiments and analyses were conducted. The first experiment is conducted in open-sky environments (including carborne, airborne, and shipborne) to evaluate the consistency between POSMind and Inertial Explorer. Additionally, experiment under urban environments is carried out to assess the performance of POSMind in realistic cases. Moreover, the practical performance of POSMind was also demonstrated with two mobile mapping cases, with the evaluation of the accuracy of point cloud. Looking forward, we plan to enhance POSMind by introducing reliable filters or optimizers, integrating observations from other sensors and utilizing the benefits of post-processing in existing powerful GNSS/SINS processing platforms. The goal is to provide a powerful GNSS/SINS post-processing service that delivers high-precision, excellent availability, and utmost reliability for diverse scenes and applications. The POSMind web and software can be freely accessed at posmind-web.com and on Kaggle website at kaggle.com/datasets/fengzhusgg/smartpnt-pos.