SYSTEM FOR SEGMENTING IMAGES OF SUBSOIL RESOURCES OF OPEN TYPE Russian patent published in 2020 - IPC G06N3/02 G06T5/50 

FIELD: physics.

SUBSTANCE: invention relates to computer engineering and can be used for segmentation of images of subsoil resources of open type. System comprises computer facilities structured on the basis of Mask R-CNN convergent neural network, which process fragments of four-channel satellite images containing images of subsoil resources of open type, with generation of segmentation algorithm of eight fragments of images at training stage by means of all possible variations of rotation of initial fragment by angle π/2, π, 3π/2 and mirror image, applying to each new obtained fragment of operators a random rotation by a small angle within [-π/12; π/12], change of scale within [1; 1.2], shift within 2 % of fragment size and chromatic distortion. As a result, a satellite image of the original size is obtained with open-cut subsoil sections allocated thereto.

EFFECT: technical result is faster processing of digital data while reducing amount of computational resources.

1 cl, 4 dwg