Abstract: This paper proposes a blurred image restoration model based on zero boundary pixels. By analyzing the causes of horizontal motion blur, it establishes matrix format of blurred model. The restoration model is perfected with conditions of zero boundary pixels to reduce cycle interference. Experimental results show that compared with Wiener filtering method, image restoration effects are better with the model when there is no noise.

Key words: zero boundary, generalized inverse matrix, blurred matrix, horizontal motion blur, image restoration, cycle interference

摘要： 提出一种基于零边界像素的模糊图像复原模型。通过分析水平运动模糊的成因，以矩阵形式建立运动模糊的一般模糊模型，利用广义逆矩阵方法对模糊图像进行复原，使用零边界像素条件完善复原模型，以减少周期干扰。实验结果表明，与维纳滤波法相比，该模型在不考虑噪声的情况下，图像复原效果更好。

关键词: 零边界, 广义逆矩阵, 模糊矩阵, 水平运动模糊, 图像复原, 周期干扰

CLC Number: 

• TN911.73