Measurement of Stain Area in Metal Surface with Particle Analysis

摘要

Measurement of the object dimension in terms of digital image processing is one of the computer vision applications. The digital image is obtained from the camera / webcam, where calibration and settings are carried out to convert the output of the camera which is a pixel of images into physical unit values (length and area). The distance between the two points on the image is known by counting the number of pixels in the image. Regression analysis method is used to get the conversion equation between variable number of pixels into units of length. The conversion equation is used to measure the length and area of 10 pieces of template cut by a CNC machine with sizes from 1x1cm to 10x10cm. The validity of the pixel calibration results in units of length was tested by measuring 10 template sizes. Measurement of the area of the stain / defect due to corrosion on metal plates is carried out on a number of samples with a random stain size, by counting the number of pixels that exist with the help of the developed LabVIEW-NI Vision application. Data acquisition and analysis systems have worked in accordance with the design concept, pixel data can be measured and converted into physical units of millimeters or centimeters, with a measurement resolution of the instrument of 0.1220 mm.