Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jul 8, 2026

Simulation of a Scaled Assembly Process with Collaboration of a Robotic Arm and Monitoring through a Vision System for Quality Control
05:47

Simulation of a Scaled Assembly Process with Collaboration of a Robotic Arm and Monitoring through a Vision System for Quality Control

Published on: August 29, 2025

Optimized image preprocessing strategies for enhanced neural network-based defect detection in industrial automation.

Sai Prakash Challa1, Melvin Alexis Lara de Leon2, Jiri Koziorek2

  • 1Department of Cybernetics and Biomedical Engineering, Faculty of Electrical Engineering, Computer Science and Cybernetics, VSB-Technical University of Ostrava, 17. Listopadu 2172/15, 70800, Ostrava, Czechia. sai.prakash.challa@vsb.cz.

Scientific Reports
|July 6, 2026
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Batteryless IoT Sensing Using Thermoelectric Energy Harvesting from Industrial Motor Waste Heat.

Sensors (Basel, Switzerland)·2026
Same author

Incremental learning approach for semantic segmentation of skin histology images.

Scientific reports·2026
Same author

Design of a new method for occupancy monitoring in smart home care with autonomous mobile robot within Internet of Things.

Scientific reports·2025
Same author

Efficient Soil Temperature Profile Estimation for Thermoelectric Powered Sensors.

Sensors (Basel, Switzerland)·2025
Same author

Constrained search space selection based optimization approach for enhanced reduced order approximation of interconnected power system models.

Scientific reports·2025
Same author

Powering Agriculture IoT Sensors Using Natural Temperature Differences Between Air and Soil: Measurement and Evaluation.

Sensors (Basel, Switzerland)·2024

This study introduces a genetic algorithm (GA) to automatically optimize image preprocessing for AI defect detection. GA-optimized methods significantly boost accuracy and reduce data needs in manufacturing.

Area of Science:

  • Computer Vision
  • Artificial Intelligence
  • Manufacturing Technology

Background:

  • Automated inspection systems in manufacturing face challenges like sensitivity to imaging variability and reliance on large datasets.
  • Current machine vision systems often require manual preprocessing, limiting accuracy and efficiency in real-world industrial settings.

Purpose of the Study:

  • To develop and evaluate a novel approach for optimizing image preprocessing strategies for neural network-based defect detection.
  • To enhance the accuracy, reliability, and data efficiency of AI-driven defect detection systems in manufacturing.

Main Methods:

  • A genetic algorithm (GA)-driven evolutionary framework was employed to systematically explore and optimize image preprocessing operations.
  • The GA evaluated a set of 48 preprocessing operations, optimizing filter sequences based on multi-objective fitness criteria including accuracy, efficiency, and robustness.
Keywords:
Deep learningDefect detectionGenetic algorithmIndustrial quality controlMachine visionManufacturing automationProduction efficiency

Related Experiment Videos

Last Updated: Jul 8, 2026

Simulation of a Scaled Assembly Process with Collaboration of a Robotic Arm and Monitoring through a Vision System for Quality Control
05:47

Simulation of a Scaled Assembly Process with Collaboration of a Robotic Arm and Monitoring through a Vision System for Quality Control

Published on: August 29, 2025

  • Experiments involved varying GA parameters such as population size and generation limits to identify optimal configurations.

    • Main Results:

    • GA-optimized preprocessing significantly outperformed raw-image baselines and manually designed pipelines across three product categories.
    • Classification accuracy improved by up to 15% with GA-optimized preprocessing.
    • Data efficiency was enhanced, requiring 30-60% fewer training images to achieve target performance.

    Conclusions:

    • Evolutionary optimization using genetic algorithms offers a robust and scalable solution for industrial defect detection.
    • The proposed method enables more reliable and efficient machine vision systems for modern manufacturing environments.
    • Automated optimization of preprocessing pipelines is crucial for overcoming limitations in current AI-based inspection systems.