storefront visibility

Inspiration

The inspiration for this project came from the challenge businesses face in selecting prime storefront locations. Visibility from the street can significantly impact foot traffic and sales, yet assessing it has traditionally been subjective and time-consuming. We wanted to create an automated, data-driven solution to simplify this process and provide actionable insights for realtors, business owners, and urban planners.

What it does

The Storefront Visibility Assessment tool evaluates how visible a storefront is from the street by leveraging the Google Maps Street View API and computer vision. Given the latitude and longitude of a location, the system captures street-level images from multiple angles, detects obstructions like trees, cars, and poles, and calculates a Visibility Score (0 to 100). A higher score indicates better visibility, helping businesses make smarter decisions when choosing storefront locations.

How we built it

Location & API Request: We collect the storefront's GPS coordinates and use the Google Maps Street View Static API to capture images from four key angles (0°, 90°, 180°, 270°). Image Processing: Using YOLOv8, a state-of-the-art object detection model, we identify storefront boundaries and obstructions (trees, cars, signs). Visibility Scoring: We calculate the percentage of visible storefront area compared to the total possible view, adjusting for angles and obstructions. Final Score: The final Visibility Score is a weighted combination of AI analysis and real-world factors like foot traffic.

Challenges we ran into

API Rate Limiting: The Google Maps API has usage limits, so we had to optimize our calls by sampling smartly across different road types and locations. Image Processing: Accurately distinguishing between storefronts and obstructions required fine-tuning the YOLO model and adjusting detection thresholds. Coordinate Conversion: Extracting precise coordinates from GIS data and converting them for API requests involved overcoming formatting inconsistencies.

Accomplishments that we're proud of

utomated Workflow: From location input to final visibility score, our pipeline runs seamlessly, providing fast and reliable results. Advanced Object Detection: Our use of YOLOv8 ensured accurate detection of obstructions and storefronts, even in complex urban environments. Efficient API Usage: By optimizing API calls and processing only necessary images, we avoided excessive costs and delays.

What we learned

API Integration: Working with the Google Maps API taught us how to optimize requests, handle rate limits, and extract useful insights from street-level data. Computer Vision: Implementing YOLOv8 deepened our understanding of real-time object detection and its applications in urban analytics. Data Processing: We learned how to clean and process location data efficiently, ensuring accurate results without unnecessary computation.

What's next for storefront visibility

Real-Time Dashboard: We plan to build a web interface where users can enter a location and instantly receive a visibility report. Historical Analysis: By integrating past Street View images, we can analyze how visibility changes across seasons and urban developments. Foot Traffic Insights: Combining visibility scores with foot traffic data will provide even more actionable insights for business owners and realtors. Expanded Coverage: We aim to scale the tool for use in cities worldwide, making it an essential resource for commercial real estate platforms.