Improving an Eggshell Inspection System With Implementation of Ultrasonic Signal and Computer Vision
Computer vision has been widely used in industries nowadays for guidance or inspection. Computer vision is a system that mimics human eyes by utilizes image taking device and processes the image. Non-destructive testing is an analysis technique to evaluate material properties without causing damage. Ultrasonic is one of the most used non-destructive testing, which utilizes pulse-waves that go through the material. Transmit time between each material differs, which will be used as reference to determine the internal properties of the material. This thesis addresses an egg inspection system which uses ultrasound to detect internal defects and computer vision to detect external defects. Ultrasonic testing will be done by sending a signal through the egg between two ultrasonic transducer. Transmit time differs with different materials placed between and will be counted with a microcontroller. Meanwhile, on the visual inspection, OpenCV is the main program that used in the software development. the steps used in this process are thresholding, edge and contour detection, filtering, and blob detection. Dirt pixels will be counted inside the algorithm, and based on the defects, the eggs will be put into their respective brackets.
B01579 | (Rack Thesis) | Available |
No other version available