Waste Sorting Vision Using Digital Watermarks on Plastic Containers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing waste sorting systems face challenges in accurately identifying and sorting plastic items, particularly those with complex surfaces, due to issues such as soiling, crumpling, and occlusion, which hinder effective code reading and sorting efficiency.
Innovation Solution
The implementation of digital watermarks on plastic items, combined with machine vision and neural networks, allows for precise identification and sorting by encoding metadata onto containers, enabling reliable sorting even under adverse conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If machine vision systems are used to read codes on plastic items, then sorting speed is improved, but accuracy deteriorates due to soiling, crumpling, and occlusion
Solution Approach 1:
The patent applies preliminary action by embedding digital watermarks and machine-readable codes into the plastic items during manufacturing, before they enter the waste stream. This ensures the identification data is already in place and protected within the item structure, allowing machine vision systems to read it accurately even after the item has been used, soiled, or damaged. The watermarks are embedded in ways that make them resistant to degradation from handling, cleaning, and environmental factors.
Solution Approach 2:
The patent uses digital watermarks as optical copies of identification data that can be read from the plastic item surface. These watermarks serve as durable copies of the item's identification information, allowing machine vision systems to access the data without direct contact with the item itself. The watermarks are designed to persist through various conditions and can be read even when the item is partially occluded or soiled, maintaining the copy's integrity while the physical item degrades.
2Reliability
If digital watermarks are embedded on complex surfaces, then identification reliability is improved, but code reading difficulty increases due to surface irregularities
Solution Approach 1:
The patent segments the identification code into multiple digital watermark components distributed across different regions of the plastic item. Rather than relying on a single continuous code that could be obscured by surface irregularities, the system divides the identification data into discrete watermark segments that can be read from multiple locations. This segmentation ensures that if one region is obscured by crumpling or soiling, other regions may still provide sufficient identification data.
Solution Approach 2:
The patent applies local quality by optimizing the watermark embedding process for different surface conditions. The system adjusts watermark characteristics, placement, and reading parameters based on the specific local conditions of each plastic item's surface. For items with complex geometries or varying surface qualities, the system adapts the watermark strategy to maximize readability in each specific location, rather than using a uniform approach across all items.
Data Source
AI summary
A waste stream is analyzed and sorted to segregate different items for recycling. Certain features of the technology improve the accuracy with which waste stream items are diverted to collection repositories. Other features concern adaptation of neural networks in accordance with context information sensed from the waste. Still other features serve to automate and simplify maintenance of machine vision systems used in waste sorting. Yet other aspects of the technology concern marking 2D machine readable code data on items having complex surfaces (e.g., food containers with integral ribbing for structural strength or juice pooling), to mitigate issues that such surfaces can introduce in code reading. Still other aspects of the technology concern prioritizing certain blocks of conveyor belt imagery for analysis. Yet other aspects of the technology concern joint use of near infrared spectroscopy, artificial intelligence, digital watermarking, and/or other techniques, for waste sorting. A variety of further features and arrangements are also detailed.