Electronics in Advanced Research Industries

7Image Vision Advances

Image vision techniques are applied mainly to detect defects in manufacturing quality processes. In this chapter different solutions are discussed oriented toward image segmentation, image classification, and image clustering for in‐line production processes. Moreover, important solutions are provided for digitalized information and data monitoring production lines, discussing some image processing architectures and industry applications. The hardware and software solutions are combined allowing big data analytics, augmented reality (AR), and predictive maintenance of machines supporting defect monitoring and quality improvement. Different solutions of sensors and implementations are matched with different scenarios in manufacturing, and in general in production processes, where autodetection of defects is fundamental for a self‐adaptive production setting. Image vision approaches are applied in different industry sectors, also involving inspection services. Some image processing approaches are discussed concerning food industries. This chapter shows pseudocodes that help to understand image processing implementation in smart systems. The discussion is mainly focused on different application fields in industries, adopting artificial intelligence (AI) solutions to improve product defect recognition. To this end, the combined use of discrete Fourier transform, K‐Means and long short‐term memory (LSTM) neural network algorithms increases the performance ...