Building Number Extraction from Street-Level Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face challenges in confirming the accuracy of business addresses when searching for street-level images, as captured images may be unclear, obstructed, or not depict the storefront clearly, leading to difficulties in identifying building numbers or postal identifiers.
Innovation Solution
A method and system that select and analyze street-level images using optical character recognition to identify and extract building numbers, comparing them to the expected building number of the address, and involving human operators for confirmation to ensure accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If street level images are captured at a high framerate, then more images are captured, but the number of images per address increases significantly making it difficult to identify the correct business storefront
Solution Approach 1:
The system performs preliminary extraction of building numbers from captured images before presenting them to users. By pre-processing the images to extract and verify address information, the system prepares the data in advance so that users receive only relevant, verified images with confirmed building numbers, eliminating the need to manually review hundreds of uncategorized images.
Solution Approach 2:
The system extracts the building number information from the image data and separates it as distinct, verifiable information. This extraction process isolates the critical address identification element from the rest of the image data, allowing the system to filter and present only those images that contain valid building numbers matching the expected address.
2Loss of information
If street level images are captured at a low framerate, then fewer images are captured, but some business storefronts may not be clearly depicted
Solution Approach 1:
The system implements a feedback mechanism where extracted building numbers are verified against expected address data. This feedback loop allows the system to confirm whether captured images actually depict the correct business storefront by checking if the extracted building number matches the expected address, thereby ensuring reliability even with reduced image capture rates.
3Speed
If automated image analysis is used to extract building numbers, then processing speed increases, but accuracy may decrease due to unclear or obstructed views
Solution Approach 1:
The patent introduces an intermediary verification step where extracted building numbers are cross-checked against expected address information from the database. This intermediary process acts as a mediator between automated extraction and final confirmation, allowing the system to maintain high processing speed while improving accuracy by filtering out incorrect extractions through database validation.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach ensures that only clear and accurate street-level images with visible building numbers are provided to users, enhancing the reliability of address confirmation and image matching.
Implementation Method 1
extracting, with the processor, an alphanumeric value from the characters within the portion of the image
Data Source
AI summary
A system and method is provided for automatically recognizing building numbers in street level images. In one aspect, a processor selects a street level image that is likely to be near an address of interest. The processor identifies those portions of the image that are visually similar to street numbers, and then extracts the numeric values of the characters displayed in such portions. If an extracted value corresponds with the building number of the address of interest such as being substantially equal to the address of interest, the extracted value and the image portion are displayed to a human operator. The human operator confirms, by looking at the image portion, whether the image portion appears to be a building number that matches the extracted value. If so, the processor stores a value that associates that building number with the street level image.