Background denoising methods in vectorization of curves
Qizhi Teng · Jisuanji gongcheng · 2005
The well log is vectorized by scanning the well log into a digital image and converting the curves in the image into digital data using the theories and technologies of image processing and recognition. One of the most important problems in the vectorization is how to remove the background and noises in the image. For this reason, the methods of frequency domain filtering, cross templet matching, linear Hough transform, intensity projection, and line segment matching for background denoising were proposed. The merits of each method was compared to the others.