Object Recognition via Image Retrieval and Tag Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current object recognition techniques require a training process that is impractical for a vast variety of objects and demands specialist expertise, leading to high costs and time consumption for improving recognition accuracy.
Innovation Solution
An object recognition device that retrieves similar images from a database using tag information associated with the images, allowing for object recognition without prior training, by employing an acquisition unit, retrieval unit, and recognition unit that utilize tag information for identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If learning-based recognition techniques are used to recognize objects, then recognition accuracy can be improved, but a training process is required which is impractical for a vast variety of objects and demands specialist expertise
Solution Approach 1:
The patent uses image retrieval to find similar images from a database instead of training a classifier. The system copies the tag information from retrieved similar images to identify objects in the target image, avoiding the need for complex training processes while maintaining recognition capability
Solution Approach 2:
The patent replaces the mechanical training process with an information-based retrieval process. Instead of mechanically training classifiers with labeled data, the system substitutes this with querying an image database for similar images and extracting tag information, eliminating the need for specialist expertise and complex training procedures
2Measurement precision
If the classifier has poor recognition accuracy, images of the learning target must be collected and used to retrain the classifier, but this retraining process requires specialist expertise that tends to be challenging for the normal user to perform appropriately
Solution Approach 1:
The system performs self-improvement by automatically retrieving similar images and extracting tag information from the database when recognition accuracy is insufficient. This eliminates the need for user intervention in the retraining process, making the system self-correcting and easy to operate for normal users
Solution Approach 2:
The system incorporates feedback mechanisms where recognition results are evaluated and used to query the image database for additional similar images and tag information. This feedback loop continuously improves recognition accuracy without requiring specialist intervention, as the system automatically adjusts based on retrieved information
3Adaptability or versatility
If a classifier is trained beforehand on multiple images of the object to be recognized, then object recognition can be performed, but creating a classifier that can be trained beforehand on the vast variety of objects that exists is impractical
Solution Approach 1:
The patent creates a universal image retrieval system that can handle any object type by querying a comprehensive image database. Instead of creating specialized classifiers for different object categories, the system uses a single retrieval mechanism that adapts to any object by finding similar images and extracting their tag information, achieving multi-functionality across diverse object types
Solution Approach 2:
The system performs preliminary action by pre-building a comprehensive image database with tag information before recognition is needed. This pre-prepared database allows the system to quickly retrieve relevant information for any object without requiring on-the-spot classifier creation, making the recognition process practical for vast object variety
Data Source
AI summary
An object recognition device includes an acquisition unit configured to acquire a recognition target image that serves as an object to be recognized; a retrieval unit configured to search an image database storing a plurality of image data in association with tag information and retrieve a similar image that matches the recognition target image; and a recognition unit configured to recognize the object included in the recognition target image on the basis of tag information associated with a similar image obtained by the retrieval unit. The recognition may select the tag information that appears most frequently among the tag information associated with the similar images as a recognition result. The recognition unit may also compute a tag information reliability score from the similar image in the retrieval result and recognize an object taking into account said reliability score.