Abstract: There is disclosed a system for automatically detecting an irregularity on a pipe. The system includes a camera arranged at an external surface of the pipe, the camera being configured to capture a Red, Green and Blue (RBG) image of a region of a pipe. One or more hardware processors are in communication with the camera and are configured to: convert the RGB image to a modified image; split the modified image into a plurality of components; generate a binary image via performing a thresholding operation which utilizes the plurality of components; and detect the irregularity on the pipe via performing a feature extraction process on the binary image. Also disclosed and described is a related method.

Abstract: There is disclosed a system for automatically detecting an irregularity on a pipe. The system includes a camera arranged at an external surface of the pipe, the camera being configured to capture a Red, Green and Blue (RBG) image of a region of a pipe. One or more hardware processors are in communication with the camera and are configured to: convert the RGB image to a modified image; split the modified image into a plurality of components; generate a binary image via performing a thresholding operation which utilizes the plurality of components; and detect the irregularity on the pipe via performing a feature extraction process on the binary image. Also disclosed and described is a related method.