Transform uploaded floor plans into 3D models using AI detection, accurate room measurements, and real-time design visualization for homeowners and renovators.
Client Overview
A home renovation and interior design company was struggling with a fundamental problem in its sales and project initiation process: the gap between what a client imagined their renovated space would look like and what the renovation team actually designed and quoted. Clients typically arrived at the initial consultation with a rough floor plan or a sketch of their existing property layout and a vague sense of what they wanted changed. Translating this into an accurate project scope, a meaningful visual representation, and a reliable cost estimate required skilled designers to spend hours manually measuring, drawing, modelling, and rendering — work that happened before any contract was signed and therefore before any revenue was generated. Designer time spent on pre-sales visualisation was significant and unpredictable. Simple projects might require two to three hours of design preparation, while complex multi-room renovations with structural changes could consume a full day or more before the client had seen anything visual. When clients chose not to proceed after seeing the initial design, this investment was entirely lost. The company's conversion rate from initial consultation to signed contract was below industry benchmarks, and internal analysis suggested that slow delivery of visual concepts was a contributing factor — clients who could not quickly visualise the outcome were making slower decisions and more frequently choosing competitors who could present concepts faster. Management wanted a platform that could dramatically compress the time from client floor plan submission to working 3D visualisation, enabling design consultants to show clients a visual model of their renovated space during or immediately after the initial consultation rather than days later.
Our Approach
Zentric Solutions built an AI-powered floor plan analysis and 3D conversion platform using computer vision and parametric 3D rendering. Clients could upload a photograph, scan, or digital file of their existing floor plan directly through a web interface. OpenCV and YOLOv8 processed the uploaded image to detect wall boundaries, doorways, window positions, room boundaries, and existing fixture placements, with a calibration step using known reference dimensions to calculate accurate real-world measurements for every detected element. Detected room structures were converted into a parametric 3D model rendered in Three.js, producing a navigable 3D representation of the existing floor plan that clients could explore in a browser without any software installation. Design consultants could then apply renovation modifications in real time — removing walls, repositioning doors, changing room configurations, and placing furniture from a curated library — with changes reflected immediately in the 3D view. The platform maintained accurate dimension tracking throughout the modification process, automatically recalculating room areas and producing updated measurement summaries as changes were applied. Generated 3D models and renovation design configurations were stored per project in the platform, allowing design consultants to continue refining concepts after the initial consultation and share updated visualisations with clients via a shareable link. The platform was built on AWS for reliable performance and integrated with the company's project management system via REST APIs to automatically create a project record with attached design files and dimension data when a client approved a concept for formal quoting. The time from client floor plan submission to initial 3D visualisation was reduced from days to under two hours in most cases.
Tech Stack
Project Tags
Everything you need to know about this project and our approach.
Zentric Solutions delivers cutting-edge digital products that streamline operations, enhance engagement, and drive lasting growth.
