Image in painting method based on evolutionary algorithm

• Main Authors: WANG, Zhaoxia, WANG, Quan, CHANG, C. S., BAI, Ming, SUN, Zhen, YANG, Ting
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2011
• Subjects: Evolutionary algorithm; Image in painting; Related contour lines; Related structure information; Texture information; Operations Research, Systems Engineering and Industrial Engineering; Theory and Algorithms
• Online Access: https://ink.library.smu.edu.sg/sis_research/5641
• Institution: Singapore Management University

This paper proposes an image in painting method based on evolutionary algorithm. The main idea of the proposed method is to restore the structure information and texture information separately. The structure information is extracted from the damaged image which is presented by several contour lines. An optimizing method based on evolutionary algorithm is proposed to analyze the structure information to find out the related structural information and calculate the related contour lines that should be connected to each other. The optimizing results as obtained by evolutionary algorithm combined with the curve fitting method gives the best matching solution to restore the dam aged structure information. The restored structure information divides the damaged image into several parts. The final in painted image as obtained after the lost texture information in every part is in painted. The efficiency of the proposed method is demonstrated by case studies in this paper.