Object Region Validation Using Dual-Criterion Image Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image recognition techniques for object detection in logistics automation, such as picking robot systems, struggle to accurately evaluate the validity of extracted object regions, leading to potential object destruction due to incorrect robot manipulations.
Innovation Solution
An information processing device that calculates an evaluation value for the validity of object regions using two distinct criteria: the degree of matching between object regions in original and augmented images, and the similarity with known object features, enhancing the accuracy of object region validation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image recognition techniques are used to extract object regions, then the system can identify objects in containers, but the validity of extracted object regions cannot be precisely evaluated, leading to potential wrong recognition
Solution Approach 1:
The patent implements a feedback mechanism by comparing the extracted object region with the original image and augmented images. The evaluation unit calculates a degree of matching between the extracted region and the original image, and between the extracted region and augmented images. This feedback loop allows the system to verify whether the extraction is valid by checking consistency across multiple image versions, thereby resolving the contradiction between extraction accuracy and validity evaluation reliability.
Solution Approach 2:
The patent uses augmented images as copies of the original image with transformed characteristics. By extracting object regions from both the original image and augmented images, and then comparing these extractions, the system creates a verification mechanism. The augmented images serve as controlled variations that help evaluate whether the extracted object region is valid, addressing the reliability issue while maintaining extraction accuracy.
2Device complexity
If single-criterion evaluation is used for object region validity, then the evaluation process is simple, but the precision and robustness of validation is insufficient
Solution Approach 1:
The patent segments the evaluation process into multiple independent evaluation criteria: (1) degree of matching between extracted region and original image, (2) degree of matching between extracted region and augmented images, and (3) similarity with known object features. Each criterion operates independently and can be evaluated separately. This segmentation allows the system to achieve high precision validation through multiple dimensions while keeping each individual evaluation component relatively simple, thus resolving the contradiction between process complexity and evaluation precision.
Data Source
AI summary
An information processing device according to an embodiment includes a hardware processor connected to a memory. The processor extracts a first object region including an object from an image. The processor calculates an evaluation value used for evaluating validity of the first object region. The evaluation value is calculated based on a first evaluation criterion and a second evaluation criterion different from the first evaluation criterion.