Context

Konkan Geoglyphs and Heritage Research Center is an organization run working towards documentation, conservation, research and tourism development for the the ancient geoglyph sites found across the Konkan Region. With over 10,000 carvings discovered across roughly 200 sites, these hold significant value as human cultural heritage. Largely undisturbed for thousands of years, they face major challenges today by accelerated human activity on the Konkan Plateau. With many of these sites are situated on private properties, the team’s efforts have spread awareness among land owners, locals and administrators around their importance as heritage, accelerating conversation efforts.

Because the heavy monsoon leaves a constant wet, reflective layer over the carvings, and vegetation overgrowth occurs during the winter, the team has to work within an extremely short 3–4 month window to clean the sites and perform archaeological documentation, which forms the basis for further research.

Documenting these shallow carvings-which are often extremely difficult to spot on the porous laterite rock and located in remote locations-make the documentation operations extremely time and resource sensitive. While, the team has uses drones to capture the photographic plans of the sites, adding much needed speed to the activity, extracting information from images into GIS software is relied on manually tracing vector outlines and adding metadata to the datasets.

Requirement

During a three-day discovery session at the organization’s Research Center in Ratnagiri, I worked alongside the team to understand their challenges, the scope of activities they conduct, and their workflows, allowing me to ideate targeted design and technology interventions. Their core needs included an efficient photogrammetry workflow to generate high-resolution 3D models and orthophotos for digital site visualization, alongside a streamlined process to accurately isolate, document, and export site layouts into GIS software.

Solution

Over 1,200 field photos were captured and processed into a dense, geocoded 3D point cloud of the Chave Dewood rhinoceros geoglyph. Based on this fieldwork, a custom photogrammetry SOP was utilized, establishing optimal camera angles and lighting balances specifically calibrated to counteract rock reflectivity and shallow carvings.

To overcome the primary technical roadblock—the labor-intensive manual tracing of rock carvings for GIS software—a computer vision solution was prototyped using Meta’s open-source Segment Anything Model 2 (SAM 2) as an accessible alternative to expensive multi-spectral imaging rigs. By manually masking a precise, custom-labeled training dataset of roughly 100 images and fine-tuning the foundation model, a specialized checkpoint was built that successfully automated the isolation of shallow geoglyph boundaries from surrounding porous rock textures.

To create a tactile souvenir for visitors, the raw 3D scan data was used to design a scaled-down, two-piece, multi-colored snap-fit 3D model of select geoglyphs. This modular approach allowed the physical pieces to snap together cleanly, cutting 3D printing time and material costs down to a fraction of standard production baselines.

Outcome

Ultimately, the engagement resulted in a fine-tuned computer vision prototype capable of automatically segmenting carvings—bypassing specialized hardware requirements and accelerating cultural heritage mapping.