Ink preset model research based on matrix singular value decomposition method with ink distribution characteristics

Abstract

The existing ink preset technology ignores the important influence of ink transfer characteristics such as ink backflow and transverse flow. This study constructs a matrix which can quantitatively express the ink distribution characteristic under corresponding graphic coverage by BP neural network. The fact is ink quantity on the substrate cannot achieve desired uniform requirements under the condition of certain graphic coverage, even if printing condition is admirable. This study uses the matrix singular value decomposition method (to obtain ink preset value by truncating singular value), so that the root mean square errors of the ink quantity on the substrate and the standard is less than the prescribed requirement. On the basis of the above research, this study proposes a new ink preset model that takes the influence of ink transverse flow into consideration. The experimental results show that the model can effectively improve the ink preset accuracy, and its application value is appreciable.