<p>Transparent exopolymer particles (TEP) are abundant gel-like colloids pivotal in marine carbon cycling and water treatment processes. Their environmental roles are governed by hierarchical architectures, yet <i>in-situ</i> structural characterization remains challenging due to transparency, fragility, and polymorphism. To address this, we developed an integrated image analysis suite combining advanced processing with statistical modeling, enabling simultaneous 2D/3D quantification of TEP morphology and intra-particle heterogeneity. This framework generates multidimensional descriptors (e.g., fractal dimensions, density gradients) for individual aggregates and assemblies. Applied to cation-mediated aggregation, it revealed divergent bridging behaviors. Mg²⁺ induced moderate size changes (2.99–4.08 µm), while Ca²⁺ drove exponential growth (1.81–187.76 µm) when ionic strength increasing from 1 to 5 mmol/L. Concurrent form factor reductions (Mg: 0.31 to 0.16; Ca: 0.44 to 0.19) quantitatively distinguish aggregation pathways. The method deciphers ion-specific assembly mechanisms and resolves subtle colloidal interactions, establishing a paradigm for colloidal system analysis with possible applications extending beyond TEP research to other subjects such as microplastic aggregation.</p>