Signal acquisition from different sources, digitization from analog signals, their analyses will be discussed. Machine vision applications, image detection and processing techniques will be covered.

You will benefit from Sabancı University’s industrial links and research strengths and you will have access to labs and facilities in SU IMC.

The course consists of theoretical components as well as applications. Teaching is provided through lectures, exercise classes, and laboratory work.

You will study to gain a thorough and detailed understanding of the principles and techniques of composite manufacturing using prepreg materials.

• Sensors;

  • Physical principles
  • Classification of Sensors: Proprioceptive/Exteroceptive, Passive/Active, Nonvisual/Visual
  • Characterizing Sensor Performance: Range, Resolution, Linearity, Bandwidth
  • In Situ Sensor Performance: Sensitivity, Accuarcy, Deterministic and Random Errors, Precision
  • Selecting Sensors for a Given Application

• Signal Processing;

  • Analog and Digital Signals
  • Sampling and Aliasing
  • Fourier Spectrum Analysis
  • FIR filters and Data Smoothing
  • Aperiodic Signals and Fourier Transform
  • FFT Algorithm o Power Spectrum

• Machine Vision;

  • Digital Images
  • Image Preprocessing (Filtering, Smoothing, Sharpening etc.)
  • Image Histograms
  • Edge and Corner Detection
  • Line and Circle Detection
  • Pattern Matching
  • Camera Calibration

• Experiments:

  Signal acquisition from different sources such as strain gauges will be performed using NI DAQ system powered by LabVIEW. Acquired signals will be digitized from analog signals while considering aliasing problem by using appropriate sampling rates. Afterwards, Fourier spectrum analysis will be performed on the digitized signals by using fast Fourier transform (FFT) in order to analyze the signals in frequency domain. Additionally, power spectrum measurement will be used to describe the energy of a signal distributed across frequency. Moreover, in order to eliminate the noise and smooth the acquired signals finite impulse response (FIR) filters will be utilized.

  Digital images (2D/3D) will be acquired by high resolution industrial cameras widely used in various machine vision applications. Afterwards, several image preprocessing techniques will be employed to enhance the image quality in MATLAB/LabVIEW environment. Histogram of images will be utilized to find out pixel intensity value distributions and will be manipulated to suit application needs such as increasing the contrast. Several edge detection methods such as Canny, Sobel, Prewitt will be applied to extract useful edge features in region of interest (ROI). Moreover, line and circle features will be extracted by Hough transform. These features can be used for different case scenarios such as segmentation, detection, and recognition. Additionally, different correlation methods will be utilized for pattern matching. Finally, camera calibration will be performed to estimate intrinsic and extrinsic parameters of the camera so that metric information can be obtained from 2D images.

This course will help you to understand the sensors, signal processing and image detections and processing techniques and machine vision applications.

This course can provide you with a strong foundation for moving into a relevant industry, improve your employability, or be the basis for further study through the pursuit of a career in research.