Photo-to-3D Garden Landscape Planner
Problem
Homeowners want to redesign their garden but have no way to turn real photos of their yard into an editable 3D model they can experiment with. Existing tools are either professional CAD (steep learning curve) or generic interior-design apps that don't handle outdoor terrain. The poster explicitly asked HN for a tool that turns garden photos into an editable 3D garden where they can drag and move things.
Opportunity
An app that photogrammetry-scans a yard from phone photos and generates an editable 3D scene with draggable plants, beds and structures, plus sun/shadow simulation. Modern photo-to-3D AI models finally make this feasible for consumers.
Market analysis
Photo-overlay and AI-render garden design is well served (iScape, Planner 5D, a wave of render-from-photo AI tools), but the specific ask, turning photos of your own yard into an editable to-scale 3D scene, is genuinely unsolved.
Market · Homeowners and DIY garden renovators, plus small landscape contractors who need proposals; iScape's scale shows proven consumer spend on outdoor visualization.
Pricing · iScape Pro anchors the category at $29.99/month or $299.99/year, but App Store reviews show consumer resistance; a realistic consumer tier is a $10 to $15 one-off or monthly sub.
Pros
- + The exact feature request is explicit and unmet in the HN thread.
- + Modern photo-to-3D models make a consumer scan-to-edit pipeline plausible for the first time.
- + Sun/shadow simulation adds real decision value that overlay apps cannot offer.
Cons
- − Gardens are close to a worst case for photogrammetry: thin structures, moving leaves, repetitive textures.
- − Making a scanned mesh or splat editable with draggable parametric objects is a very hard UX problem.
- − Incumbents with distribution (iScape, Planner 5D) can bolt generative AI onto what they already have.
Existing / similar tools
- → iScape ↗
- → Planner 5D ↗
- → GardenBox 3D ↗
- → Neighborbrite
Source
Hacker News (Ask HN)
The trap here is conflating two products: an AI re-render of your photo (solved, commoditized, cheap) and an editable digital twin of your actual terrain (open, hard). The second one lives or dies on geometry quality, and vegetation is notoriously bad input for reconstruction, so the realistic wedge is scanning only the ground plane, boundaries and fixed structures, then placing parametric plants on top of that simplified terrain. That hybrid would already beat iScape for anyone whose question is “will this tree shade the patio in July” rather than “make my yard look pretty”, and landscape pros paying for proposals are a far better first customer than one-off homeowners.