Advanced Vision System for Tire Cord Fabric Analysis
Budget: $250 – $750 USD
We are seeking experienced professionals or serious companies to develop an advanced vision system for analyzing tire cord fabric images captured by a line scan camera at a speed of 100 meters per minute. The project involves the development of algorithms and tools for defect identification, width measurement, and mesh measurement of the fabric.
Milestones:
Defect Identification and Width Measurement:
Develop an algorithm capable of learning the characteristics of good fabric and identifying any anomalies or defects without generating false positives or negatives.
Create a tool to accurately measure the width of the fabric in each captured image, with a measurement accuracy within ±5mm.
Mesh Measurement:
Implement a method to count the number of warp and weft threads within a given area of the fabric, ensuring accurate and adaptable measurements for different fabric densities.
Deliverables:
Defect Identification Algorithm: A robust algorithm that can differentiate between natural variations and actual defects in the fabric.
Width Measurement Tool: A tool that measures the width of the fabric accurately, adhering to the specified tolerance of ±5mm.
Mesh Measurement Implementation: A reliable method for counting warp and weft threads, adaptable to various fabric densities.
Software Environment and Requirements:
Software: Halcon (MVTec), Deeplearning techniques (object detection, segmentation, anomaly detection)
Operating System: Ubuntu 20.04
Programming Language: C++
Performance: The total cycle time for defect detection, width measurement, and mesh measurement should be less than 50 to 70 ms on the specified hardware (12fps at 4096x1024 resolution).
Image Information:
Image Size: 4096 x 1024, tiled into 4 images of 1024 x 1024 each.
Image Type: Mono8/Color8
Hardware Information:
Graphics Card: GTX 1660 Ti
CPU: Intel i5, 16GB RAM, 1TB Hard Disk
Note: Please only contact if you have proven experience and expertise in this domain. The dataset and additional documentation are provided via the link below:
Dataset and Documentation
Only serious companies and professionals should reach out. We look forward to collaborating with those who can deliver high-quality results within the specified requirements.
Data set: https://drive.google.com/file/d/1TJcDgemiHzf5-Q0s6G8S_7sr7FZxQAHl/view?usp=sharing
Milestones:
Defect Identification and Width Measurement:
Develop an algorithm capable of learning the characteristics of good fabric and identifying any anomalies or defects without generating false positives or negatives.
Create a tool to accurately measure the width of the fabric in each captured image, with a measurement accuracy within ±5mm.
Mesh Measurement:
Implement a method to count the number of warp and weft threads within a given area of the fabric, ensuring accurate and adaptable measurements for different fabric densities.
Deliverables:
Defect Identification Algorithm: A robust algorithm that can differentiate between natural variations and actual defects in the fabric.
Width Measurement Tool: A tool that measures the width of the fabric accurately, adhering to the specified tolerance of ±5mm.
Mesh Measurement Implementation: A reliable method for counting warp and weft threads, adaptable to various fabric densities.
Software Environment and Requirements:
Software: Halcon (MVTec), Deeplearning techniques (object detection, segmentation, anomaly detection)
Operating System: Ubuntu 20.04
Programming Language: C++
Performance: The total cycle time for defect detection, width measurement, and mesh measurement should be less than 50 to 70 ms on the specified hardware (12fps at 4096x1024 resolution).
Image Information:
Image Size: 4096 x 1024, tiled into 4 images of 1024 x 1024 each.
Image Type: Mono8/Color8
Hardware Information:
Graphics Card: GTX 1660 Ti
CPU: Intel i5, 16GB RAM, 1TB Hard Disk
Note: Please only contact if you have proven experience and expertise in this domain. The dataset and additional documentation are provided via the link below:
Dataset and Documentation
Only serious companies and professionals should reach out. We look forward to collaborating with those who can deliver high-quality results within the specified requirements.
Data set: https://drive.google.com/file/d/1TJcDgemiHzf5-Q0s6G8S_7sr7FZxQAHl/view?usp=sharing