摘要
文中针对大数量的规则图像碎片进行了拼接复原研究,在图像碎片缺少外形轮廓这一匹配特征和碎片数量庞大的前提下,提出了一种基于最小生成树原理的规则图像碎片快速复原算法。通过计算图像碎片边缘像素差异值的方法对碎片进行匹配,再运用贪心策略的思想,通过最小生成树原理对图像进行复原框架设计,完成了对规则图像碎片的快速复原。而且相比现有算法,文中算法无需知道原始图像的尺寸,更为符合实际应用情况。仿真结果表明,文中算法完成了对大数量图像碎片的复原工作,具有快速、准确的特点。
The matching and restoration for large number of square image pieces are studied in this paper. On the premise of lacking outline and large quantity of pieces,a fast restoration algorithm for square image pieces based on Minimum Spanning Tree (MST) is put forward. It calculates the difference of pixel value on the edge to match two pieces,then with the idea of the greedy strategy,the structure for restoration of square image pieces is designed,completing the quick restoration of square image pieces finally. Compared with the existing algorithms ,this algorithm does not need to know the size of the original image,more accorded with the actual application situation. Simu- lation indicates that it can complete the restoration for tens of thousands pieces rapidly and accurately.
出处
《计算机技术与发展》
2016年第6期69-72,77,共5页
Computer Technology and Development
基金
国家自然科学基金资助项目(61272043)
重庆市基础与前沿研究重点项目(cstc2013jj B40009)
重庆科技研发基地能力提升项目(cstc2014pt-sy40003)
关键词
规则碎片
匹配
复原
最小生成树
square pieces
matching
restoration
minimum spanning tree