Efficient Reconstruction of Industrial Images Using Optimized HMK Splines

Abstract

In the context of the fourth industrial revolution (also referred to as Industry 4.0), industrial image processing forms a valuable basis for newly conceived applications and services, especially image reconstruction as one of the key technologies of industrial image processing plays an important role in industrial engineering. To develop an effective method for industrial image reconstruction, this article proposes the optimized hierarchical many-knot (HMK) splines (abbreviated as OHMK splines) method. In the scheme, a fine merged control lattice is derived from the hierarchy of coarse-to-fine control lattices, and then the final reconstruction function obtained by OHMK splines is the function generated on the merged fine control lattice, which is not obtained by superimposing all functions generated on the coarse-to-fine control lattices. Therefore, the computation for the final reconstruction function relies on the number of control points in the merged fine control lattice, not depending on all the control points in the hierarchy. Through the experimental results, we can find that the good reconstruction performance for the given image can be obtained by using the OHMK splines method, and the computational overhead of the OHMK splines method is superior to that of the HMK splines method.