Updated: Jan 18, 2022
Inspecting indented surface or surface roughness is not an easy task if only using thresholding technique because difference lighting condition changes the inspection result drastically and making the system not stable. Therefore, In this example, we demonstrates how to detect small texture defects on surface using Fast Fourier Transformation (FFT).
Fourier Transformation is used to analyze the frequency characteristic of various filters. For images; 2D Discrete Fourier Transformation (DFT) is used to find the frequency domain. A fast algorithm called FFT is used for calculation of DFT. More intuitively, for a sinusoidal signal, if the amplitude varies so fast in short time, you can say it is a high frequency signal. If it varies slowly, it is a low frequency signal. You can extend the same idea to images, where at the edge points, or noises are high frequency contents. if there is no much changes in amplitude, it is a low frequency component.
First, we construct a high pass filter of gaussian. Then, the images and the filter are convolved by using FFT. Finally, the defects are detected in the filtered images by using morphology operators.
So, in Squidify System, simply adding these tools to an inspection sequence, and link up operators inputs and output connections, we can easily resolve complex mathematic solutions for images processing.
In a group of images, yielded results as shown below
In such inspection operations, it takes approximately 40ms for a 445 pixels (width) x 303 pixels (height) image and a longer processing time to be expected for a larger image pixels size. However, it is still considered efficient by applying FFT.