Abstract

Product quality control is currently the leading trend in industrial production. It is heading towards the exact analysis of each product before reaching the end customer. Every stage of production control is of particular importance in the food and pharmaceutical industries, where, apart from visual issues, additional safety regulations are demanded. Many production processes can be controlled completely contactless through the use of machine vision cameras and advanced image processing techniques. The most dynamically growing sector of image analysis methods are solutions based on deep neural networks. Their major advantages are fast performance, robustness, and the fact that they can be exploited even in complicated classification problems. However, the use of machine learning methods on high-performance production lines may be limited by inference time or, in the case of multiformated production lines, training time. The article presents a novel data preprocessing (or calibration) method. It uses prior knowledge about the optical system, which enables the use of the lightweight Convolutional Neural Network (CNN) model for product quality control of polyethylene terephthalate (PET) bottle caps. The combination of preprocessing with the lightweight CNN model resulted in at least a five-fold reduction in prediction and training time compared to the lighter standard models tested on ImageNet, without loss of accuracy.

Highlights

  • Food and drink manufacturers are obliged to control the quality of their products

  • Unlike general neural networks models that can be used for many different tasks [15], the presented approach of designing of the classification model requires a priori knowledge of the optical system and/or inspected object

  • The article presents the application of deep neural network methods for the detection and classification of defects in closures of liquid packaging on production lines

Read more

Summary

Introduction

Food and drink manufacturers are obliged to control the quality of their products (regulatory issues). A frequently used solution is to multiply cameras observing the object from different sides to register the entire surface of the object [7,8] It may make the measurement system more complicated and the development of algorithms for automatic image analysis may be much more difficult as the views from different cameras may differ from each other. The applicability of such systems in the same form and with the same image processing algorithms on a different production line would be limited. The analysis of several images of the same object increases the requirements for computing power, which, as a consequence, may limit the production efficiency (which is unacceptable from the manufacturer’s point of view) or significantly affect the cost of the system itself

Objectives
Methods
Findings
Discussion
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call