大功率LED宽幅数码图像冲印的优化与仿真

Optimization and Simulation of High Power LED Width Digital Image Printing

大功率LED宽幅数码图像的分辨率较低,像素之间无明显特征差。利用传统算法进行LED图像分割的过程中,由于像素的灰度值和平面坐标值相近,导致出现像素分割后每个方格之间极易出现断续边界、色彩失真、降低冲印质量等问题。提出基于大功率LED宽幅数码图像冲印的图像分割优化方法。首先对待冲印图像中的有效像素进行提取,计算出有效像素的平面坐标值和灰度值,然后进行加权处理,得出像素的形心坐标。最后通过像素变化产生的形心坐标值的变化,得出待冲印图像亚像素级别的图像识别结果,完成对待冲印图像的优化。实验结果表明,利用改进算法进行图像分割,有效地提高了冲印的精度。

An optimization method for image segmentation was proposed based on high power LED width digital image printing. Firstly, the effective pixels of the needed printing image were extracted, to calculate the plane coor- dinate values and the gray values of the effective pixels. Then, weighted processing was implemented to conclude the centroid coordinates of pixels. Finally, through the change of centroid coordinate after pixel variations, the image rec- ognition results of the needed printing image in sub - pixel level were obtained, so as to complete the optimization of needed printing image. The experimental results show that using the improved algorithm for image segmentation, can effectively improve the accuracy of the printing.