The problem of optimal cutting of honeycomb blocks is considered in the article. The urgency of using irregular shapes application cutting from honeycomb blocks made of modern composite materials is substantiated. The problem is to obtain a cut of a given shape from honeycomb blocks. The optimal cutting plane of the single cell face of a honeycomb block is found on the basis of calculations for the stability and stretching force in adjacent faces applying a finite element method. The problem of obtaining faces and nodal points of a honeycomb block when calculating the coordinates of the optimal cut points of a real honeycomb block using machine vision methods. The most suitable algorithms, a part of the OpenCV library, are considered. The authors decided to develop their own algorithm for recognizing an irregular cellular structure since the existed algorithms did not solve the stated problem. As opposed to existing algorithms, the proposed one finds all the nodal points and faces even on the deformed sections of the honeycomb block avoiding false faces. The method helps to find coordinates of all nodal points and determine the connection of adjacent points with a face. As a result, real faces, not perfect geometrically, are replaced by line segments. The test result proves its correctness and practical applicability. The proposed method solves an optimal cutting problem of aluminum honeycomb block. It should be noted that the proposed method is customizable and can be applied to any honeycomb or cellular panels.INDEX TERMS Image recognition, composite materials, cutting tools, machine vision.