Control Freak 2 is a multi-platform input system that supports multi-touch, accelerometer, keyboard, mouse and game controllers. Make your game multi-platform without touching a single line of code! (In 99% of cases no code modification is needed). Main features : * Advanced touch control system that can perfectly emulate traditional input sources (digital, analog, delta, scroll and even low-level touches). On-screen controls can be composed out of button, joysticks, steering wheels, track pads and super touch zones. State of touch controls can be bound to virtual keys, named buttons and axes. Super Touch Zone, the most advanced control type of the pack, lets you bind every finger gesture imaginable - for example, single-finger tap can act as a right mouse button click, swiping horizontally with three fingers can emulate mouse scroll wheel and two-finger twist gesture can be bound to an analog axis that rotates the camera! * Simple API - there's no need to learn new API or rebuild you code to work with event-based input system... Familiar Input-like static class will make you feel right at home. * Script Converter will automatically convert your project's scripts! * With Control Freak Assistant you will be able to quickly create custom Input Rigs, complete with touch controls, with just a couple of clicks.

C#图片对比，支持图片局布和整体图片比对，原图采样和目标图原图采样进行比对，特征匹配 在特征描述中，我们得到了512bit的二进制描述符，该描述符的列是高方差——>低方差的排列，而高方差表征了模糊信息，低方差表征了细节信息，与人眼视网膜相似，人眼先处理的是模糊信息，再处理细节信息。因此，选取前128bit即16bytes进行匹配（异或），若两个待匹配的特征点前16bytes距离小于设定的阈值，则再用剩余的位信息进行匹配。这种方法可以剔除掉90%的不相关匹配点。注意：这里的16bytes的选取是建立在并行处理技术（SIMD）上的，并行处理技术处理16bytes与处理1bytes的时间相同；也就是说，16bytes并不是固定的，如果你的并行处理技术能处理32bytes与处理1bytes的时间相同的话，那么你也可以选取前32bytes。

环境是opencv2.4.9，将已有的一些提取方式进行了总结，对一些参数也进行了标注，部分算法的参数含义并未标注，但将其默认参数和默认值进行了标注，共11种提取方式，未使用simpleblob.

基于OpenCV2.4.2和VS2008平台下的ORB和FREAK局部特征描述子的对比实验代码

基于ORB-FREAK算法的移动增强现实系统实现，牛玉涛，李红波，针对移动设备上现有注册算法存在运算量大、实时性差的问题，提出了一种基ORB-FREAK算法的移动增强现实三维注册技术。为缩减算法时间

CVPR2012上freak算法的源代码。

怪胎发现者 python 脚本来查找容易受到 Freak Attack 攻击的主机

FREAK feature the MatLab mex实现

特征点检测与匹配，采用freak算子，是一个简单的应用