Inverted Index Object Recognition Using PostScript Shape Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional object recognition methods using PostScript data face high processing loads when dealing with large amounts of data, making it time-consuming to identify and associate objects in captured images.
Innovation Solution
The method involves generating an inverted index based on PostScript data, allowing for the extraction of object shapes and text information, and using a processor to identify similar objects by comparing the generated text information with stored identification information, thus reducing the need for extensive object drawing and comparison processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If PostScript data is used to represent object shapes, then object recognition can be performed, but the processing load increases significantly when the amount of data is large
Solution Approach 1:
The patent extracts only the essential shape information from PostScript data by generating simplified text information that represents the drawing indications. This extraction process removes unnecessary complexity while retaining the core geometric features needed for object recognition, thereby reducing processing load without sacrificing recognition accuracy.
Solution Approach 2:
Instead of directly comparing complex PostScript data representations, the patent inverts the approach by converting shapes into text information and using this text representation for comparison. This inversion transforms the problem from comparing complex drawing instructions to comparing simplified textual descriptions, significantly improving processing efficiency.
2Reliability
If conventional object recognition methods are used with large amounts of PostScript data, then comprehensive object analysis is possible, but it takes much time to draw and compare shapes
Solution Approach 1:
The patent performs preliminary conversion of object shapes into text information before the actual recognition process. By pre-processing the shape data into a standardized text format that captures essential drawing indications, the system prepares the data in advance for rapid comparison, eliminating the need for time-consuming real-time drawing operations during object recognition.
Solution Approach 2:
The patent creates simplified text-based copies of the original PostScript shape representations. These text copies retain the essential geometric information needed for identification while being much faster to process and compare. The system uses these textual copies as intermediaries to perform rapid object matching without repeatedly processing the full PostScript data.
Data Source
AI summary
A file generating device (100) extracts, when a captured image captured by an image capturing device is acquired, based on the acquired captured image, a shape of an object included in the captured image and generates, based on the extracted shape, text information that includes a drawing indication of the shape. The information processing apparatus (200) refers to a storage unit that stores therein, regarding each of a plurality of objects, identification information on the objects in association with the text information that includes a drawing indication of the objects and acquires identification information on an object, from among the plurality of objects, that is associated with the text information in which similarity to the generated text information satisfies a criterion.