Affine-invariant recognition of handwritten characters via accelerated KL divergence minimization
Affine-invariant recognition of handwritten characters via accelerated KL divergence minimization
复制标题
通过加速 KL 散度最小化的仿射不变识别手写字符
DOI：
10.1109/icdar.2011.221
复制
发表时间：
2011
期刊：
影响因子：
--
通讯作者：
Y.Yamashita
中科院分区：
文献类型：
--
作者：
T.Wakahara;Y.Yamashita
本文提出了一种新的，仿射不变的图像匹配技术，通过加速KL（Kullback-Leibler）的分歧最小化。首先，我们通过将像素值的总和设置为1，将图像表示为概率分布。其次，我们使用高斯核密度估计将仿射参数引入到两幅图像的概率分布中。最后，我们确定最佳的仿射参数，最小化KL分歧，通过迭代方法。特别是，不使用传统的非线性优化技术，如Levenberg-Marquardt方法，我们设计了一个加速迭代方法，通过有效的线性近似适应KL发散最小化问题。在手写体数字库IPTP CDROM 1B上的识别实验表明，该方法在抑制计算量的前提下，识别率达到91.5%，比基于正常KL散度的简单图像匹配方法的识别率83.7%高得多。
This paper proposes a new, affine-invariant image matching technique via accelerated KL (Kullback-Leibler) divergence minimization. First, we represent an image as a probability distribution by setting the sum of pixel values at one. Second, we introduce affine parameters into either of the two images' probability distributions using the Gaussian kernel density estimation. Finally, we determine optimal affine parameters that minimize KL divergence via an iterative method. In particular, without using such conventional nonlinear optimization techniques as the Levenberg-Marquardt method we devise an accelerated iterative method adapted to the KL divergence minimization problem through effective linear approximation. Recognition experiments using the handwritten numeral database IPTP CDROM1B show that the proposed method achieves a much higher recognition rate of 91.5% at suppressed computational cost than that of 83.7% obtained by a simple image matching method based on a normal KL divergence.