Attribute-Specific Identifier Selection for Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing inspection apparatuses face a trade-off between identification accuracy and versatility when detecting objects with various attributes, as they are designed to determine apparent quality quickly rather than achieving high accuracy for objects with diverse attributes.
Innovation Solution
An identifying apparatus that selects an identifier specific to the attribute of an object, using a storage unit to store multiple identifiers trained for different attributes, an imaging unit to capture images, a specifying unit to determine the object's attribute, and a selecting unit to choose the appropriate identifier for processing, thereby reducing processing load and memory usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple identifiers are simultaneously executed to detect detection targets in objects with various attributes, then identification accuracy for diverse objects is improved, but processing load and memory usage increase
Solution Approach 1:
The patent segments the identification process by dividing multiple identifiers into separate functional units. Each identifier is stored in the storage unit and can be independently selected and executed based on the specific detection task, rather than simultaneously executing all identifiers. This segmentation reduces the processing load while maintaining the capability to achieve high identification accuracy for various object attributes.
2Measurement precision
If multiple identifiers are simultaneously executed to detect detection targets in objects with various attributes, then identification accuracy for diverse objects is improved, but memory usage increases
Solution Approach 1:
The patent implements preliminary action by pre-storing multiple identifiers in the storage unit during the system setup phase. These identifiers are prepared in advance and associated with specific object attributes or detection scenarios. When a detection task is executed, the system selects and loads only the relevant identifier from the stored set, rather than maintaining all identifiers in active memory simultaneously. This approach reduces memory usage during operation while preserving identification accuracy.
3Productivity
If a single identifier is used to detect detection targets in objects with various attributes, then processing load is reduced, but identification accuracy for specific object attributes decreases
Solution Approach 1:
The patent implements dynamics by making the identifier selection adaptive and dynamic based on the input object's attributes. The system includes a selection unit that dynamically chooses the appropriate identifier from the stored set based on characteristics of the detection target. This dynamic selection mechanism allows the system to maintain high processing efficiency by executing only one identifier at a time, while simultaneously achieving high identification accuracy by selecting the most suitable identifier for each specific detection task.
Data Source
AI summary
An identifying apparatus may include: an imaging unit configured to store multiple identifiers that are trained to identify, from images of objects, the presence or absence of a detection target included in the objects, the identifiers being stored in association with attributes of the objects, and to capture an image of a predetermined object; a specifying unit configured to specify an attribute of the predetermined object; a selecting unit configured to select, from the multiple identifiers, a first identifier stored in association with the specified attribute; an input unit configured to input the image of the predetermined object to the first identifier; and an output unit configured to output the presence or absence of the detection target included in the predetermined object, which is output from the first identifier.