Mail Handling Device State Detection Using Dual Learning Algorithms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing monitoring systems for mail handling devices rely on light barriers and cameras, which are inefficient and prone to human error. Light barriers can only detect issues at specific points and are easily blocked, while continuous camera monitoring requires significant manual labor and is susceptible to misinterpretation.
Innovation Solution
A computer-implemented method using a first learning algorithm to extract features from an image data stream of a mail handling device, and a second learning algorithm to determine the state of the device based on these features, allowing for reliable and efficient detection of issues such as jams without the need for calibrated cameras or extensive manual monitoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If light barriers are used to monitor handling devices, then detection at specific positions is achieved, but the system is easily blocked and cannot provide continuous monitoring
Solution Approach 1:
The patent replaces mechanical light barriers with a camera-based visual inspection system. The camera captures images of the handling device and mailpieces, and image processing algorithms automatically detect jams and anomalies. This substitution eliminates the blocking issue of light barriers while maintaining detection accuracy through computational analysis of visual data.
Solution Approach 2:
The patent introduces an intermediary image processing system between the camera and the detection output. This intermediary layer processes visual data through algorithms that can infer jam conditions from camera images, providing reliable detection without direct contact with the mailstream, thus avoiding blocking issues.
2Reliability
If cameras are used to monitor handling devices, then continuous monitoring is possible, but significant manual labor and human error are required
Solution Approach 1:
The patent implements self-service by using automated image processing algorithms to analyze camera feeds. The system automatically detects jams, anomalies, and handling device states without requiring human intervention. The algorithm processes images independently, providing continuous monitoring with minimal operational complexity.
Solution Approach 2:
The patent replaces manual visual inspection with automated computer vision systems. Image processing algorithms automatically analyze camera data to detect handling device states, eliminating the need for continuous human monitoring while maintaining reliable and continuous detection.
3Measurement precision
If calibrated cameras are used for monitoring, then accurate measurement is achieved, but the system complexity and cost increase
Solution Approach 1:
The patent changes the approach from physical calibration to computational calibration. Instead of mechanically calibrating camera positions and angles, the system uses image processing algorithms to infer spatial relationships and detect anomalies. This parameter change maintains measurement accuracy while significantly reducing system complexity.
Solution Approach 2:
The patent introduces an intermediary computational layer that processes camera data without requiring direct physical calibration. The image processing algorithms act as an intermediary, translating raw camera data into meaningful detections while abstracting away the complexity of calibrated camera positioning.
4Measurement precision
If photoelectric sensors are used, then isolated stream detection is possible, but dense parcel streams block sensors and require dense networks
Solution Approach 1:
The patent replaces physical photoelectric sensors with a camera-based system. The camera captures images of the entire handling area, and image processing algorithms detect jams and anomalies. This substitution eliminates the need for dense sensor networks while maintaining detection capability, as the camera provides comprehensive visual coverage of the entire stream.
Solution Approach 2:
The patent transitions from point-based sensor detection to area-based image analysis. Instead of using multiple point sensors arranged in a dense network, the system uses a single camera to capture two-dimensional images of the entire handling area, providing comprehensive detection with reduced complexity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A computer-implemented method for determining a state (10) of a handling device (1) for mailpieces (2) is provided. The method comprises a) determining or obtaining an image data stream (4) of the handling device (1) when handling mailpieces (2), b) generating at least one image packet (5, 6) from the image data stream (4), wherein the image packet (5) comprises a plurality of individual images (7), c) determining features (8) of each image (7) of an image packet (5, 6) by means of a first learning algorithm (9), and determining the state (10) of the handling device (1) based on the features (8) of each image (7) of an image packet (5, 6) by means of a second learning algorithm (11), wherein the first learning algorithm (9) differs from the second learning algorithm (11).Furthermore, a method for training or retraining a first learning algorithm (9) and/or a second learning algorithm (11) for determining a state (10) of a handling device (1) for mail items (2), a computer program and a monitoring system for determining a state (10) of a handling device (1) for handling mail items (2) are provided.