Quick Tutorial

Please note that is just a "general case" quick tutorial that will work with a reasonably high number of images, although not always with 100% success. For the goriest details, read the FAQ section or study the Hough transform section.

Open a *.BMP, *.JPG or other supported image from the LOAD button. Go to the Original tab. If necessary, use the Min , Max , Negative or Median buttons to filter or correct the image (e.g. filter out noise, make it "White on Black" etc. ). Select an edge detection method from the "Edge Detection" menu. "Prewitt" is the default one. Press the "Edges" button in the Original tab to perform edge detection on your image. You will be taken to the EdgeDTC tab, with the edges of your image visible. Perform Binarization of your image. Generally, pressing the Otsu Binary button will be enough. You will be taken to the Binary tab, with your edge image binarized.
Select the minimum and maximum DIAMETER you want the program to detect from the"Min.Diameter" and "Max.Diameter" panels.

Select your calculation method, e.g. "Circular Hough Transform" , by pressing the "Use CHT" key. Wait until calculations are finished. Please note that they can take quite a long time to complete when working on large images, images with a lot of shapes and when you want to detect a broad spectrum of circular diameters.
Use the "View Diameter" control in the "Transform" tab to view the different Hough Transform spaces for each examined circle diameter. If the images are too faint , try using the "Amplify" trackbar to increment their brightness.
Finally, press the "Detect Circles" key to see the detected circles superimposed to your original image.
You can now...uhm...admire your image, save it, copy to clipboard or view it in a new window. You can repeat the procedure for the same image, or another one, with different parameters, to see how the final result changes. Try experimenting with different transformation methods, tolerance values and diameter ranges to achieve better results.