Automatic detection method for optical structure of metallurgical coke

The invention discloses an automatic detection method for an optical structure of metallurgical coke, which comprises the following steps of: shooting a group of 200-time images with two different polarization analysis angles in the same vision field under the conditions of orthogonal polarization of an optical microscope and addition of a lambda complementary color plate, and identifying a background, a silk group and a fragment structure in the vision field; identifying the anisotropic optical tissue through judgment and calculation of data of two pictures with different polarization analysis angles in the same area, calculating shape characteristic parameters of the identified anisotropic tissue color-changing particles, performing classification through the shape characteristic parameters, and finally calculating the proportion of different types of color-changing particles, and obtaining the proportions of isotropic optical tissues and other optical tissues, and repeating the steps until the proportions