Video Object Identification via Reverse Image Search and OCR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Consumer measurement companies face challenges in identifying logos and brands in various media formats due to the complexity and cost of existing image recognition techniques.
Innovation Solution
The method involves extracting video frames and using a reverse image search engine to identify objects, where textual information from search results pages is captured and processed using optical character recognition, allowing for parsing-presentation independent object detection without needing to handle dynamic data formats like HTML or JSON.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional image recognition techniques are used to identify logos and brands in video, then identification capability is achieved, but processing complexity and cost increase significantly
Solution Approach 1:
The patent introduces a reverse image search engine as an intermediary component between video frame extraction and object identification. Instead of using complex traditional image recognition systems directly, the system extracts video frames, uses them as queries to a reverse image search engine, and processes the returned search results to identify objects. This intermediary approach simplifies the overall system architecture while maintaining identification accuracy.
Solution Approach 2:
The system captures screenshots of search results pages returned by the reverse image search engine and uses optical character recognition (OCR) to extract textual information from these visual copies. This copying approach allows the system to process information in a standardized image format rather than dealing with dynamic HTML or JSON data structures, reducing processing complexity.
2Loss of information
If dynamic data formats like HTML or JSON are processed directly from search results, then complete information is obtained, but parsing complexity and operational difficulty increase
Solution Approach 1:
Instead of parsing dynamic HTML or JSON data formats directly from search results, the system captures visual copies (screenshots) of the search results pages. These static image copies can be processed using standardized OCR techniques, eliminating the need for complex parsers that must handle various dynamic data formats. This approach maintains information completeness while dramatically simplifying the processing operations.
Solution Approach 2:
The patent replaces the mechanical parsing process (which would require writing and maintaining complex code to interpret HTML/JSON structures) with an optical recognition system. OCR technology automatically extracts textual information from the visual representation of search results, substituting a more robust and easier-to-operate mechanism that is not sensitive to changes in data format structures.
Data Source
AI summary
Example systems disclosed herein are to process image frames with a reverse image search engine to generate corresponding search results pages, capture screenshots corresponding respectively to the search results pages, generate base query records corresponding to respective ones of the screenshots that have respective textual information that matches a base search term, the respective ones of the base query records including the base search term and at least portions of the respective textual information from the corresponding screenshots, determine an object search term based on a frequency analysis of the textual information included in the base query records, generate object query records corresponding respectively to ones of the base query records that have respective textual information matching the object search term, and identify a first object depicted in the video based on at least one criteria applied to the object query records.