提出，对于BMP图象，直接在象素值上提取特征； 对于JPEG图象，在其DCT系数中提取特征。
Proposed that features be extracted from spacial pixel values for BMP image and from DCT coefficients for JPEG images.
在读取信号之前和之后，你可以提取存在于象素中的噪音信号，然后在后续的过程中通过减去这些噪音信号来达到消除它们的目的。
Before and after reading out the signals, you can extract noise existing within the pixels themselves, and then cancel it out by subtracting it at a later stage.
针对模糊图像的弱点，分析了通过增强象素模糊属性对比度来提取边界特征的基本原理，实现了两幅图像的配准。
Aiming at the disadvantage of fuzzy image, by making some analysis about extracting edge feature based on fuzzy contrast enhancement algorithm, two images registration is achieved.
提出以计算象素点的方向游程长度为基础，完全基于汉字的结构知识的笔划提取算法。
Put forward the calculation that pixel direction distance length as foundation, the stroke extraction way according to the structure knowledge of Chinese characters completely.
该方法首先基于邻域灰度极值提取边界候选图像，然后以边界候选象素及其邻域象素的二值模式作为样本集，输入边缘检测神经网络进行训练。
The method uses a logical judgment algorithm to get edge candidate images, and then edge pixels and their neighbor pixels compose the binary samples of the BP neural network.
该方法首先基于邻域灰度极值提取边界候选图像，然后以边界候选象素及其邻域象素的二值模式作为样本集，输入边缘检测神经网络进行训练。
The method uses a logical judgment algorithm to get edge candidate images, and then edge pixels and their neighbor pixels compose the binary samples of the BP neural network.
应用推荐