Development of computer vision for inspection of bolt using convolutional neural network
| Autor: | Kandasamy Jayakrishna, T.T.M. Kannan, T. Vignesh, J. Chandradass, A. John Rajan |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
010302 applied physics
Computer science business.industry Machine vision 02 engineering and technology General Medicine 021001 nanoscience & nanotechnology 01 natural sciences Mechanical components Interchangeability Convolutional neural network Raspberry pi Development (topology) 0103 physical sciences Quality check Computer vision Artificial intelligence 0210 nano-technology business |
| Zdroj: | Materials Today: Proceedings. 45:6931-6935 |
| ISSN: | 2214-7853 |
| Popis: | The inspection of bolt is difficult in conventional quality check procedure. Computer vision inspection is a suitable method to find interchangeability. The aim of the present study is to develop a device to detect defects in the bolt with the help of computer vision technology. Many traditional techniques are used to find the defects in mechanical components using computer vision in Industries. This paper focuses the development of vision system for measurement and inspection of bolt using camera attached with algorithms. This work is mainly built on the self-learning convolutional neural network to implement computer vision technology to detect the defects. The algorithm is built on the C language and tested repeatedly. After that algorithm is impended on the raspberry pi board, and a neutral stick is attached to the raspberry pi model to operate the algorithm. The camera is attached with the raspberry pi model to capture the image, analyze and identify the defects of bolt. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect