This dataset identifies 1,307 patent families published between 2015 and mid-2026 that combine core super-resolution techniques, such as single-molecule localization and stimulated emission depletion, with practical implementation constraints like localization precision, drift correction and labelling density. The true global figure will be higher, since any keyword-based search has boundaries, but 1,307 gives a workable sense of scale. Because publication lags filing by roughly 18 months, the 2025-2026 counts in particular will rise as more applications publish.

The ranking covers 100 assignees, with the leader holding 152 records against 40 at fifth place and 24 at tenth — a steep drop after the top filer rather than a gradual taper. The top five assignees combined account for 27.5% of all 1,307 records, and the top ten for 38.8%, so a majority of the filing activity sits outside this concentrated group. The assignee base mixes academic institutions, microscope and instrument makers, and diagnostics companies.

Filing volume rose from 90 records in 2017 to a peak of 120 in 2020, then declined to 87 by 2024 — a 27% drop over that three-year span, which is the last period with reasonably complete publication data. This looks more like a maturing core technology than a shrinking field: the foundational optical and reconstruction methods are heavily claimed, pushing new filings toward specific applications. Figures for 2025 onward are still incomplete due to the normal filing-to-publication lag.

Relative to the dense claim coverage in material analysis (G01N, 57.2% of records) and optical systems (G02B, 31.5%), areas such as drift correction algorithms, photoswitching buffer chemistry, labelling density optimisation for dense samples, and bioinformatics pipelines for localization data (G16B, 6.3%) carry noticeably lighter claim density. These are reasonable starting points for scouting first-claim opportunities, though each requires its own prior-art check rather than relying on IPC share alone.

US11676247B2, held by Institut Pasteur and published in 2023, covers reconstructing a synthetic dense super-resolution image from diffraction-limited image sequences using a two-stage process: a standard single-molecule localization reconstruction followed by an artificial neural network step. It blocks that specific combination of SMLM localization output feeding into a neural network for dense image synthesis. It does not, on its face, block other reconstruction approaches that skip the neural network step or use different input combinations, so a design-around is plausible but should be checked claim-by-claim.