Filter Inspection Device for Flavor Element Embedding Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for inspecting filter cigarettes fail to accurately determine the quality and proper embedding of string-like flavor containing elements within columnar filters attached to cigarette ends, including whether the element is embedded and its location.
Innovation Solution
A filter inspection method and apparatus that photographs the filter end face, uses pattern matching to find the center position and approximate circle of the filter, detects the flavor element area by color difference, and determines the distance between the element and the filter center to assess the element's quality and embedding accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing inspection methods photograph the filter end face and analyze color components, then dust or dirt adherence can be detected, but the quality and proper embedding of string-like flavor containing elements cannot be accurately determined
Solution Approach 1:
The inspection process is segmented into distinct functional modules: center position finding (via approximate circle fitting), flavor element area detection (via color component analysis), distance calculation, and quality determination. This segmentation allows each module to be optimized independently, improving measurement precision while managing complexity through modular design.
Solution Approach 2:
The patent introduces an intermediary coordinate system and mathematical transformations to bridge the gap between the photographed image data and the physical embedding quality. By establishing a relationship between the filter center position, element position, and acceptable embedding tolerances, the system can accurately determine flavor element quality without direct physical measurement.
2Loss of information
If the inspection system analyzes color components to detect dust or dirt, then surface contamination can be identified, but the location and embedding quality of flavor containing elements cannot be determined
Solution Approach 1:
The inspection system performs multiple functions using a single imaging setup: it detects dust/dirt adherence through color component analysis, identifies the filter center position via approximate circle fitting, locates flavor containing elements by color difference, and determines embedding quality through distance calculation. This multi-functionality reduces information loss while avoiding the need for separate inspection devices for each function.
Solution Approach 2:
The system changes analysis parameters based on inspection objectives: using color component thresholds for dust detection, circle fitting algorithms for center positioning, and distance calculations for embedding quality assessment. By dynamically adjusting which parameters are analyzed and how they are processed, the system comprehensively captures all relevant information from a single image without requiring complex additional hardware.
3Manufacturing precision
If the flavor containing element is embedded along the axis of the filter, then proper alignment is achieved, but inspection accuracy is affected by filter distortion or positioning variations
Solution Approach 1:
The system performs preliminary actions to establish reference frames before measuring embedding accuracy: it first detects the filter end face轮廓, fits an approximate circle to determine the filter center position, and establishes a coordinate system. These preliminary steps create a distortion-compensated reference framework that enables accurate measurement of flavor element embedding position even when the filter is distorted or mispositioned in the image.
Solution Approach 2:
The patent handles the asymmetric nature of filter distortion by using approximate circle fitting rather than assuming perfect circular symmetry. The method accommodates elliptical or irregular filter shapes by finding the best-fit circle parameters (center coordinates and radius), thereby compensating for asymmetric distortions and maintaining measurement precision in determining whether the flavor element is properly embedded along the filter axis.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An approximate circle approximated to the profile of a filter end face in an inspection image obtained by photographing the filter end face of a filter cigarette is obtained, and the center position of the approximate circle is found. An area indicating a string-like flavor containing element embedded in the filter is detected in the inspection image, and the position of gravity center of the area is found. Distance between the gravity center position indicating the flavor containing element and the center position of the approximate circle is determined, and thus, the quality of the flavor containing element embedded in the filter is inspected.