Research On Stochastically Seeded Clustered Digital Halftoning Method Of Laser Printer

The laser printer has become the important output device of the image information in desktop office system as for its fast speed,high quality and low working noise.The laser printer needs to transform the continuous digital image into a non-continuous image firstly,which is named as digital halftoning technology.The imaging mechanism of the laser printer determines the inevitability of the nonlinear distortion,which makes it easier for the printout image to have dot gain and dot loss.Therefore,it is difficult to ideally reproduce the tone of the input image.The error diffusion digital halftoning method has a good reconstruction of the image details,and the particle sensation produced in the human eye is small.However,the error diffusion digital halftoning method has the direction lag,and there is a regular texture in the bright area of the image.This paper chooses the stochastically seeded clustered digital halftoning method to halftone the input image.First of all,the blue noise digital halftoning method is used to generate stochastically seeded points.Then,we use the growing method to build the Delaunay triangular mesh.Subsequently,the threshold matrix is generated according to the location of the pixels in the triangle.In order to improve the number of the different thresholds in matrix,we use Baer table to optimize the initial threshold matrix.Finally,the adjusted threshold matrix is used to get the digital halftoned image.The digital halftoned images show that the stochastically seeded clustered digital halftoning method can solve the problem that the bright area of the image has regular texture in error diffusion digital halftoning method.In order to verify the nonlinear distortion performance of the stochastically seeded clustered digital halftoning method,the linear evaluation method of tone reproduction based on Munsell lightness is used to evaluate the quality of the printout image.The green noise digital halftoning method has good stability for nonlinear distortion of printer.As a result,this digital halftoning method is selected as the reference method to ensure the validity of the verification test.We use the stochastically seeded clustered digital halftoning method,the error diffusion digital halftoning method and the green noise digital halftoning method to get digital halftoned images.And then evaluate the quality of the printout images.The data shows that the stochastically seeded clustered digital halftoning method is better to resist the nonlinear distortion of the laser printer in the bright area of the image,and it can reproduce the tone of the image in a more linear way.