Capsule Seam Image Acquisition Using HSV Brightness Channel
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting splitting-in-insertion defects in capsules, especially when the capsule body and cap have the same color, are inefficient and prone to missed detections due to low contrast and unobvious features, leading to inaccurate and time-consuming manual or machine vision processes.
Innovation Solution
A method and apparatus for acquiring a capsule seam image by converting the image into a preset channel with step changes, performing highlighting and noise reduction processes to enhance the capsule seam feature, and compensating to obtain a complete outline of the capsule seam, allowing for accurate defect detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If machine vision detects capsules with same-color body and cap, then detection automation is improved, but detection precision deteriorates due to low contrast and unobvious seam features
Solution Approach 1:
The patent converts the capsule image from RGB color space to HSV color space and extracts the V (value/brightness) channel to enhance the visibility of the capsule seam. This color space transformation and channel selection effectively highlights the seam features even when the capsule body and cap have the same color, resolving the low contrast problem while maintaining automated detection
Solution Approach 2:
The patent transforms the 2D image data into a 1D projection profile along the vertical direction, converting spatial information into a dimensional representation that emphasizes the seam's step-change characteristics. This dimensional transformation enables precise seam detection by focusing on the brightness variation pattern rather than relying on color contrast
2Productivity
If conventional machine vision methods are used for same-color capsules, then detection speed is improved, but missed detection rate increases due to inability to identify seam features
Solution Approach 1:
By transforming to HSV color space and using the V channel, the method enhances seam visibility without requiring multiple shots or complex color analysis, maintaining high detection speed while significantly reducing missed detection rate for same-color capsules
Solution Approach 2:
The patent performs preliminary image preprocessing including noise reduction filtering and gradient calculation before seam detection. These preliminary actions enhance the seam features in advance, enabling reliable detection in a single shot and improving both speed and reliability
3Measurement precision
If manual detection with sieve plate is used, then detection accuracy is improved for size-based defects, but processing efficiency deteriorates due to mechanical structure limitations
Solution Approach 1:
The patent replaces the mechanical sieve plate system with an image processing-based detection method. By using computer vision to analyze capsule images and detect seam features through brightness profile analysis, the system achieves both high accuracy and high efficiency, eliminating the mechanical speed limitation while maintaining defect detection capability
4Reliability
If capsule size tolerance range is considered in detection, then false positive rate is reduced, but detection precision for actual defects deteriorates due to overlapping size ranges
Solution Approach 1:
The patent extracts and focuses specifically on the capsule seam region by projecting the V channel image data vertically and analyzing the brightness profile. This extraction of the seam-specific feature allows detection to be independent of overall capsule size variations, enabling accurate defect detection while maintaining low false positive rate
Solution Approach 2:
The patent applies local quality analysis by examining the brightness gradient characteristics specifically at the seam location rather than analyzing the entire capsule. This localized approach to detecting the step-change pattern at the seam enables precise defect identification without being affected by global size tolerances
Data Source
AI summary
A method for acquiring a capsule seam image includes: acquiring a first image component, where the first image component includes an image component of a capsule seam in a preset channel; in the first image component, the image component of the capsule seam in the preset channel has step change; performing a highlighting process on the first image component according to an insertion direction from the capsule cap to the capsule body to obtain a second image component, where the second image component at least includes an enhanced capsule seam feature; and performing noise reduction and compensation on the second image component to obtain a target image component, where the target image component includes a complete outline of the capsule seam.