Metal nanoparticles (NPs) exhibit multiple unique technological characteristics compared to their bulk or micrometer counterparts, such as coalescing or sintering at lower temperatures. This enables low-cost manufacturing of electronic products by using NPs as nano pastes. Three-dimensional (3D) observation of their dynamic sintering process is an emerging approach to deepen our scientific and technological understanding of the different pathways underlying sintering phenomena at the nanoscale. However, quantitatively evaluating the details of NP ensembles, including measurements of neck length, relative geometrical features of the NPs, initial dispersion, and their relative positions, from 3D data remains labor-intense and error prone so far due to the non-uniformity of both NP shape and their distribution, as well as the large volume of data. Here, we developed an empirical-parameter-free scheme for the quantitative evaluation of microstructural parameters in 3D that characterize neck growth and densification. The developed scheme was based only on a few relatively simple mathematical assumptions, such as the center point of any geometry being the farthest from the outside of the geometry, and the neck being a flat plane formed between NPs. This paper demonstrates the successful extraction of both the neck and the connectivity of NP ensembles from in situ scanning transmission electron microscopy (STEM) tomography data, providing a quantitative description of NPs’ coalescing/sintering behavior.