A 2D linear canonical transform partition algorithm(二维线性正则变换剖分算法)

Abstract

Due to the flexible selection in parameter values of the two-dimensional linear canonical transform, different forms of linear canonical transformation series can be obtained. To enhance the computation speed of linear canonical transform image representation, in this paper, an effective image representation algorithm is proposed by using linear canonical transform series combined with a non-uniform partition algorithm. It selects appropriate expression to approximate regional image pixels by utilizing the flexibility of multi-parameter linear canonical transform. The partition rectangle number in each 16×16 grid forms the feature matrix which is then combined with Arnold transform to establish a lightweight image protection method that can resist different forms of attacks. Experimental results show that the proposed method has a particular function to detect attacks.(二维线性正则变换可灵活选择参数值，得到不同形式的线性正则变换级数。为提高线性正则变换图像表示算法的计算速度，将线性正则变换级数与非均匀剖分算法相结合，得到了一种有效的图像表示算法。利用多参数线性正则变换的灵活性，选择合适的表达式逼近区域图像像素。图像的特征矩阵由16×16网格的剖分矩形数构成，将其与Arnold变换相结合构建了能抵抗不同形式攻击的轻量级图像保护算法。实验结果表明，算法具有检测图像是否遭受攻击的能力。)

Keywords

• 2d linear canonical transform(二维线形正则变换)
• non-uniform partition(非均匀剖分)
• image protection(图像保护)