Saw Chain Sharpness Detection Using Neural Image Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for determining the sharpness of a saw chain in motor chain saws are inefficient and impractical, as they require precise and reproducible conditions for temperature and image analysis, making it difficult to quickly and reliably assess whether the saw chain is sharp or dull.

Innovation Solution

A method using an artificial neural network to evaluate roof cutting images, allowing for quick and accurate assignment of saw chain sharpness regardless of recording conditions, with a high success rate even at lower image resolutions, enabling the use of inexpensive imaging devices and reducing storage space.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional image analysis methods are used to determine saw chain sharpness, then measurement precision can be achieved, but the method requires absolutely reproducible conditions (lighting, background, zoom) which makes it impractical for quick field assessment

Engineering Contradiction:
Improvesharpness determination accuracyVSAvoidrecording condition requirements
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The patent replaces traditional manual image analysis methods with an artificial neural network system. The neural network automatically processes roof cutting edge images and determines sharpness without requiring manual measurement or controlled recording conditions, thus substituting mechanical/manual operations with an intelligent automated system that is both precise and easy to operate.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the approach from measuring physical dimensions (width of roof cutting edge) to analyzing pixel color values and patterns in the image. By transforming the measurement parameters from geometric dimensions to color-based features that the neural network can process, the system achieves accurate sharpness determination without requiring controlled lighting or background conditions.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If high resolution photos are taken to capture the small roof cutting edge, then measurement precision improves, but the file size increases storage requirements and processing time

Engineering Contradiction:
Improveroof cutting edge detection accuracyVSAvoidimage data volume
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent extracts only the essential information needed for sharpness determination - the color values and patterns of pixels along the roof cutting edge - rather than storing and processing the entire high-resolution image. The neural network processes these extracted features to determine sharpness, significantly reducing data volume while maintaining measurement precision.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent uses a resolution that is sufficient but not excessive for the task. Instead of using the maximum possible resolution, the system uses just enough resolution to capture the necessary pixel color information for the neural network to accurately determine sharpness, avoiding the storage and processing overhead of unnecessarily high resolution images.

Inventive Principle:
Principle #16Partial or excessive action

3Ease of operation

If temperature measurement is used to assess saw chain sharpness, then a non-contact method is provided, but the measurement is influenced by moisture content of the wood being cut

Engineering Contradiction:
Improvesharpness assessment methodVSAvoidsharpness determination accuracy
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The patent uses color changes in the roof cutting edge as detected by the camera's image sensor to determine sharpness. Different sharpness states produce different color patterns and pixel values in the captured image, which the neural network analyzes to assess chain condition. This color-based approach is not influenced by wood moisture content, providing reliable measurements.

Inventive Principle:
Principle #32Color changes

4Measurement precision

If manual evaluation of roof cutting edge width is performed, then measurement precision can be achieved, but the process is time-consuming and requires laboratory conditions

Engineering Contradiction:
Improvesharpness measurement accuracyVSAvoidsharpness assessment speed
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces manual measurement and evaluation processes with an artificial neural network that automatically analyzes roof cutting edge images. The neural network rapidly processes pixel color values and patterns to determine sharpness, eliminating the time-consuming manual measurement process while maintaining or improving measurement precision, and eliminating the need for laboratory conditions.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP4167180A1Method for determining the sharpness of a saw chain of a motorised chain saw
Publication Date: 2023.04.19 ANDREAS STIHL AG & CO KG
  • EP4167180A1 patent drawingFigure 1~2
  • EP4167180A1 patent drawingFigure 3~4
  • EP4167180A1 patent drawingFigure 5~6

AI summary

The invention relates to a method for determining the sharpness of a saw chain (1) of a motor chainsaw (2). The saw chain (1) comprises at least one cutting link (3) with a top edge (4). An imaging device (5) captures an image of the top edge (4) of the cutting link (3). An evaluation unit (6), which includes an artificial neural network, performs an evaluation of the top edge image. In the evaluation unit (6), a sharp state and a dull state of the saw chain (1) are defined. Based on the top edge image, the saw chain (1) is assigned to a sharp state or a dull state using the artificial neural network.