cloudinautumn的博客

Hive编程指南（中文版）,作者：Hive O'REILLY，2013年12月第一版 非影印版，文字十分清晰

HoshiKata用C#写的一个卡尔曼滤波实例代码，下面是原文描述： Kalman filters allow you to filter out noise and combine different measurements to compute an answer. They are a particularly powerful type of filter, and mathematically elegant. Without a matrix math package, they are typically hard to compute, examples of simple filters and a general case with a simple matrix package is included in the source code. Developed by Rudolf Kalman and others as an ideal way to estimate something by measuring something, its vague applicability (estimate something by measuring something) is where it is so powerful, there doesn’t need to be a direct relationship between measured and estimate.

sift算法详解及应用，非常适合初学者阅读

提出一种新的基于无限冲激响应( IIR) 低通滤波的Retinex 图像增强算法, 该算法具有边缘保留功能, 不会产生光晕效应, 而且计算量较小. 并将该算法与实数编码遗传算法相结合, 使其具有针对不同图像特点自适应选择参数的功能. 通过对低照度图像和雾天图像的实验表明, 该算法能够根据图像的特点自动选择出较合理的参数, 处理后的图像在清晰度和色彩上较MSR 及MSRCR 算法具有更好的效果.