The invention discloses a method for quantitatively evaluating the surface pulverization roughness of a composite insulator. An image matrix sequence is read through the visible light image processingtechnology and a matlab software editing algorithm, and the effective feature parameters of the image are extracted, so as to achieve the precise judgment of the surface aging features of a compositeinsulator of a power transmission line. The method mainly comprises the steps: extracting a visible light image of the composite insulator, and carrying out the quantitative representation of the surface roughness of the composite insulator; employing an unmanned plane for shooting, and carrying out the detection of a to-be-detected composite insulator under the live-line condition; carrying outthe onsite collection of the visible light image of the to-be-detected composite insulator, extracting the feature parameters of the image, and precisely evaluating the surface roughness of the imagethrough the comparison of the roughness with the known surface roughness.