LED Waste Sorting via Vision Recognition and Coordinate Calculation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current recycling systems for end-of-life LED lighting lack effective classification and pre-treatment technologies, leading to environmental issues due to inadequate separation of LED-mixed waste, which results in the mixing of fine powders with metal components, making it difficult to sort and classify bulb-type LED lighting from other forms of LED lighting.
Innovation Solution
A classification device and system utilizing vision recognition through a conveyor system with cameras, an analysis part using image classification algorithms, and a computation unit to calculate coordinates for precise sorting and separation of bulb-type LED lighting from LED-mixed waste, allowing for automated classification and conveyance to a classifying conveyor.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If uniform crushing is applied to all LED waste lighting without classification, then processing efficiency is improved, but sorting accuracy deteriorates because fine powders mix with metal components making classification difficult
Solution Approach 1:
The system performs preliminary classification of LED waste lighting into different types (bulb-type, tube-type, panel-type) before crushing using vision recognition and coordinate calculation. This preliminary sorting action separates different housing forms while maintaining their structural integrity, preventing fine powder contamination of metal components that would occur with uniform crushing, thus resolving the contradiction between processing efficiency and sorting accuracy.
2Measurement precision
If vision recognition with coordinate calculation is implemented for classification, then sorting accuracy is improved, but device complexity increases due to multiple components including cameras, analysis parts, and conveying mechanisms
Solution Approach 1:
The vision recognition system integrates multiple functions into a unified classification device: image capture by cameras, image analysis and recognition, coordinate position calculation, and coordinated conveying to different sorting destinations. This multi-functional integration achieves high sorting accuracy for different LED housing forms while managing system complexity through functional consolidation rather than separate independent systems.
3Ease of operation
If automated classification system is implemented, then labor cost is reduced, but initial investment increases due to advanced vision recognition equipment and automated conveying systems
Solution Approach 1:
The automated classification system enables self-service operation where the vision recognition device automatically identifies LED waste lighting types, calculates coordinates, and directs conveying without human intervention. The system processes waste lighting autonomously through image capture, analysis, coordinate determination, and automated sorting, significantly reducing labor costs while the initial investment is offset by long-term operational efficiency and accuracy improvements.
Data Source
AI summary
A classification device and classification system for sorting and classifying bulb-type LED lighting through vision recognition from LED-mixed waste lighting having different forms of housing are provided. The classification device including: an input conveyor configured to move a plurality of LED-mixed waste lighting; a sensing part configured to obtain images of the LED-mixed waste lighting moving along a path of the input conveyor using a plurality of cameras; an analysis part configured to search for bulb-type LED lighting from the images obtained from the sensing part using an algorithm for image classification and recognition; a computation unit configured to calculate X, Y, and Z coordinate positions of an object classified as bulb-type LED lighting in the analysis part; and a conveying part configured to pick up and convey the object with the X, Y and Z coordinates calculated by the computation unit to a classifying conveyor.