Blog

, , ,

A talk on Machine Vision Systems

Department: Electronics and Communication Engineering

Name of the event: A talk on Machine Vision Systems

Date of the event: 26/08/19

Objective of the event: To discuss applications of Machine Vision Systems and the implementation methods with emphasis on image processing and machine learning techniques

 Speaker: Prof. Vivek Kanhangad, Department of Electrical Engineering, IIT Indore

 Details of participants: Students of ECE II and III year.

Total no. of participants: 40

Highlights of the talk:

Department of ECE conducted an expert talk on Machine Vision Systems for II and III year ECE students on 26/08/19. Prof. Vivek Kanhangad from IIT Indore was the speaker. Basics of image processing and its applications in Machine Vision Systems were discussed. The discussion was broadly divided in three categories of Human Vision System, Applications of Machine Vision Systems and the methods to implement Machine Vision Systems. The talk started with the application of industrial inspection for quality control. The methods for sensing technologies, data acquisition data analysis and dimensionality of signals were discussed. The image processing techniques such as chemical imaging, X-ray imaging, infrared-imaging and multispectral imaging were discussed. The role of these techniques in textile industries, electronics industries and optical applications such as contact lens inspection were discussed. Several industrial examples involving pattern recognition, image and signal analysis and machine learning were taken. Several places where manual inspection is difficult are identified and the methods to substitute the manual inspection by automated intelligent solutions were discussed. Different processing stages of pattern recognition systems such as sensing, preprocessing, face/feature extraction and decision making were discussed. Difference between speech recognition and speaker recognition was discussed. A brief example of face recognition system that involves image segmentation, feature vector extraction and classification using support vector machines, decision tree and Gaussian Mixture Models was discussed. At the end an industrial example of salmon and sesa fish classification using a robotic arm was taken. The talk concluded with a question answer session.