Ctrl2Cap v2.0

最新推荐文章于 2023-05-11 17:56:26 发布
原创 最新推荐文章于 2023-05-11 17:56:26 发布 · 1.5k 阅读 · 0 ·
CC 4.0 BY-SA版权
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
文章标签：

#keyboard #class #windows #filter #ddk #file

Ctrl2Cap是一款内核模式设备驱动程序，用于将系统键盘类驱动中的大写锁定字符转换为控制字符。适用于从UNIX迁移到NT系统的用户，提供在标准PC键盘上将控制键替换为大写锁定键的功能。适用于NT 4.0及更高版本，包括64位Windows XP、Windows Server 2003、Vista和Longhorn Server。安装通过运行"ctrl2cap /install"命令，卸载则使用"ctrl2cap /uninstall"。

By Mark Russinovich

Published: November 1, 2006

Introduction

Ctrl2cap is a kernel-mode device driver that filters the system's keyboard class driver in order to convert caps-lock characters into control characters. People like myself that migrated to NT from UNIX are used to having the control key located where the caps-lock key is on the standard PC keyboard, so a utility like this is essential for our editing well-being.

Ctrl2cap works on NT 4.0 and higher, as well as 64-bit Windows XP, Window Server 2003, Vista and Longhorn Server.

Installation and Use

Install Ctrl2cap running the command "ctrl2cap /install" from the directory into which you've unzipped the Ctrl2cap files. To uninstall type "ctrl2cap /uninstall".

Building Ctrl2cap

Ctrl2cap requires ntddkbd.h from the ddk/src/input/inc directory, so you must first copy that file to the ctrl2cap directory. To build a standard NT 4 driver type "buildnt4" and to build a Win2K WDM driver type "buildnt5". Output files are placed in the /release subdirectory.

How Ctrl2cap Works

On NT 4 Ctrlcap is actually quite trivial. It simply attaches itself to the keyboard class driver so that it will catch keyboard read requests. For each request, it posts an I/O completion callback, at which point it takes a peek at the scancode that is being returned. If it happens to be a caps-lock, ctrl2cap changes it into a left-control.

On Win2K Ctrl2cap is a WDM filter driver that layers in the keyboard class device's stack above the keyboard class device. This is in contrast to the Win2K DDK's kbfiltr example that layers itself between the i8042 port device and the keyboard class device. I chose to layer on top of the keyboard class device for several reasons:

•

It means that the Ctrl2cap IRP_MJ_READ interception and manipulation code is shared between the NT 4 and Win2K versions.

•

I don't need to supply an INF file and have the user go through the Device Manager to install Ctrl2cap - I simply modify the appropriate Registry value (the keyboard class devices's HKLM/System/CurrentControlSet/Control/Class UpperFilters value).

The disadvantage of my approach is (and this an advantage or disadvantage depending on your point of view):

•

Because I don't install with an INF file via the Device Manager, the user is not warned that the Ctrl2cap driver file is not digitally signed by Microsoft.

In this particular case I felt that the advantages outweigh the disadvantages. However, before you model a Win2K keyboard filter on Ctrl2cap I strongly suggest that you study the kbfiltr example from the Win2K DDK. Kbfiltr's interception point in the key input sequence makes it very easy for kbfiltr to inject keystrokes into the input stream.

More Information

For more information on writing filter drivers (drivers that attach themselves to other drivers so that they can see their input and/or output), here are sources to check out:

•

The Windows NT and Windows 2000 DDK sample /src/storage/filter/diskperf

•

The Windows 2000 DDK sample /src/input/kbfiltr

•

"Examining the Windows NT File System," by Mark Russinovich and Bryce Cogswell, Dr. Dobb's Journal, February 1997 the accompanying file system filter driver, Filemon

Download Ctrl2cap (48 KB)

Yuan-2.0M32 推理脚本 与vllm加速
强化学习曾小健
07-11 2082
这个脚本通过torchrun启动了一个分布式推理服务器，用于运行 "Yuan-2.1B" 模型。脚本设置了各种模型参数、分布式训练配置以及推理服务器的相关配置。脚本中的关键配置包括指定使用的 GPU 设备、设置 NCCL 超时时间、定义GPT 模型参数、指定 Tokenizer 类型和路径、以及加载模型检查点等。
Ctrl2Cap
知道的越多，不知道的越多
06-16 194
Ctrl2Cap是一个将CapsLock键映射为Ctrl键的实用工具。
Ctrl2cap Keyboard Filter
02-23
Ctrl2cap Keyboard FilterExample keyboard filter driver for NT4 and Win2KCopyright (C) 1996-1999 Mark Russinovichhttp://www.sysinternals.comTo install: - Run "ctrl2cap.exe /install" from install directory - RebootTo uninstall: - Run "ctrl2cap.exe /uninstall" from install directory - RebootTo build source: - Copy ddksrcinputincntddkbd.h to sys directory - Run buildnt4.bat under NT 4 DDK to build NT 4 driver * symbols placed under symbols.nt4 * driver produced as releasectrl2cap.nt4.sys - Run buildnt5.bat under Win2K DDK to build Win2K driver * symbols placed under symbols.nt5 * driver produced as releasectrl2cap.nt5.sys
[内核编程] 键盘过滤第一个例子ctrl2cap（4.1~4.4）汇总，测试
weixin_30632899的博客
11-02 223
键盘过滤第一个例子ctrl2cap（4.1~4.4）汇总，测试 完整源代码 /// /// @file ctrl2cap.c /// @author wowocock /// @date 2009-1-27 /// #include <wdm.h> #include <ntddkbd.h> #includ...
《windows内核安全与驱动开发》ctrl2cap中的ObReferenceObjectByName疑问
weixin_30920853的博客
04-20 266
　　国内有关于windows内核驱动这块的书籍实在是甚少，不过好在《windows内核安全与驱动开发》这本书还算不错（内容方面），但是不得不说这本书在许多地方存在着一些细节上的问题。比如我今天要谈的这个话题。 　　在这本书的键盘过滤这个章节，作者对ObReferenceObjectByName个函数的用法做了介绍，并指明这是一个非文档化函数（可以用，但是在MSDN的文档中以及SDK的头文件中没有...
《寒江独钓》ctrl2cap问题代码
08-21
《寒江独钓》一书中键盘过滤驱动ctrl2cap疑似有问题的代码及蓝屏截图
V4L2系列 之 V4L2驱动框架
小嵌同学的博客
04-23 5201
本篇文章主要介绍Linux内核中的V4L2框架，本篇文章所用内核版本：linux-4.19v4L2 (Video for Linux 2），是linux的一套视频框架，共主体位于内核，可以理解为是整个linux系统上面的视频源捕获驱动框架。其广泛应用在嵌入式设备、移动端以及个人电脑设备上面，市而上使用视频图像采集的设备如：手机、IPC、行车记录仪都会用到这个框架来进行视频采集。v4L2允许 应用程序 控制图像传感器以及传输格式，应用程序 借此完成拍照、预览、视频记录等图像传感器数据应用。
das 2.0开发记录
wherwh的专栏
09-23 1494
本文目前是工作过程记录的素材整理，尚 1.内存对象定义及访问 核心的内存对象描述对应的系统资源，主要包括机构，用户，服务器。 原来本地和平台是2套分别定义的。 本地： MEM_SVR_DATA定义:common\APBase.h SERVERINFO,USERINFO,ORGINFO定义:    common\APBase.h     统一定义到:     .i_m
PD2.0流程具体分析
最新发布
09-05
set_bit(PD_CTRL_REG, ENABLE_PD2_EXTENSIONS); // 配置回调函数 register_interrupt(PD_MSG_RECEIVED, msg_handler); } // 消息处理回调 void msg_handler(uint8_t msg_type) { switch(msg_type) { case PD...
Windows资源监控工具汇总
liuying263的专栏
02-22 3285
在利用LoadRunner进行性能测试的时候，Windows服务器的资源是经常需要监控的对象。其实除了LoadRunner提供的计数器，似乎Window服务器并不像Unix或者Linux提供众多的性能监控方法，比如Top或者vmstat等。另外有很多第三方的工具可以选用以增强Windows服务器的监控途径，下面是一些工具的列表。 【监视类】 CPUMon v2.0 CPU性能监视工具。可以获取CPU计数器信息。该版本集成了Perfmon。http://www.sysinternals.com/ntw2k
import numpy as np import cv2 as cv import random as rng from collections import deque import board_demo.ros_robot_controller_sdk as rrc import os import time from Fuzzy_pid import FuzzyPID # 导入模糊PID类 # 解决Qt平台插件问题 os.environ["QT_QPA_PLATFORM"] = "xcb" # 初始化硬件控制板 board = rrc.Board() board.enable_reception() print("硬件控制板初始化完成") # 设置随机种子 rng.seed(12345) # 舵机控制类 class ServoController: def __init__(self): self.current_pos = [500, 500] self.horizontal_range = (100, 900) self.vertical_range = (200, 800) self.laser_offset_x = 0 self.laser_offset_y = 0 print(f"舵机控制初始化: 水平范围={self.horizontal_range}, 垂直范围={self.vertical_range}") print(f"激光偏移量: X={self.laser_offset_x}, Y={self.laser_offset_y}") def set_position(self, horizontal, vertical, move_time=500): target_h = int(horizontal + self.laser_offset_x) target_v = int(vertical + self.laser_offset_y) print(f"\n[舵机设置] 原始目标: H={horizontal:.2f}, V={vertical:.2f}") target_h = max(self.horizontal_range[0], min(self.horizontal_range[1], target_h)) target_v = max(self.vertical_range[0], min(self.vertical_range[1], target_v)) if target_h != horizontal + self.laser_offset_x or target_v != vertical + self.laser_offset_y: print(f"位置限制生效: H={target_h}, V={target_v}") self.current_pos = [target_h, target_v] board.bus_servo_set_position(0.5, [[1, target_h], [2, target_v]]) print(f"设置舵机位置: H={target_h}, V={target_v}, 移动时间={move_time}ms") return target_h, target_v def get_position(self): return self.current_pos def main(): # 打开摄像头 cap = cv.VideoCapture(0) if not cap.isOpened(): print("无法打开摄像头") return # 设置分辨率 cap.set(cv.CAP_PROP_FRAME_WIDTH, 640) cap.set(cv.CAP_PROP_FRAME_HEIGHT, 480) print(f"摄像头设置: 宽度={640}, 高度={480}") # 创建边缘检测对象 ed = cv.ximgproc.createEdgeDrawing() EDParams = cv.ximgproc_EdgeDrawing_Params() EDParams.MinPathLength = 120 EDParams.PFmode = True EDParams.MinLineLength = 30 EDParams.NFAValidation = True ed.setParams(EDParams) print(f"边缘检测参数: MinPathLength={EDParams.MinPathLength}, MinLineLength={EDParams.MinLineLength}") center_queue = deque(maxlen=100) cv.namedWindow("Detected Ellipses", cv.WINDOW_NORMAL) # 初始化模糊PID控制器 - 分别用于X轴和Y轴 # 水平控制 (X轴) fuzzy_pid_x = FuzzyPID( base_kp=0.2, kp_min=0.01, kp_max=3, base_ki=0.000, ki_min=0.000, ki_max=0.1, base_kd=1, kd_min=0.0, kd_max=10, output_min=-10, output_max=10, e_min=-316, e_max=316, # 假设图像宽度640，中心偏差范围±320 ec_min=-100, ec_max=100 ) # 垂直控制 (Y轴) fuzzy_pid_y = FuzzyPID( base_kp=0.5, kp_min=0.1, kp_max=2.0, base_ki=0.000, ki_min=0.000, ki_max=0.02, base_kd=0.1, kd_min=0.0, kd_max=0.5, output_min=-10, output_max=10, e_min=-240, e_max=240, # 假设图像高度480，中心偏差范围±240 ec_min=-10, ec_max=10 ) servo = ServoController() # 获取图像中心 ret, frame = cap.read() if not ret: print("无法读取摄像头图像") return image_height, image_width = frame.shape[:2] center_x, center_y = image_width // 2, image_height // 2 print(f"图像中心: X={center_x}, Y={center_y}") # 帧率计算变量 frame_count = 0 start_time = time.time() while True: ret, frame = cap.read() if not ret: print("无法读取摄像头图像") break frame_count += 1 # 图像处理 gray = cv.cvtColor(frame, cv.COLOR_BGR2GRAY) blurred = cv.GaussianBlur(gray, (5, 5), 0) ed.detectEdges(blurred) ellipses = ed.detectEllipses() # 绘制图像中心点 cv.circle(frame, (center_x, center_y), 5, (0, 255, 255), -1) cv.line(frame, (center_x - 15, center_y), (center_x + 15, center_y), (0, 255, 255), 1) cv.line(frame, (center_x, center_y - 15), (center_x, center_y + 15), (0, 255, 255), 1) cv.putText(frame, "X", (image_width - 20, center_y - 10), cv.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 255), 1) cv.putText(frame, "Y", (center_x + 10, 20), cv.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 255), 1) ellipse_detected = False if ellipses is not None: print(f"\n检测到 {len(ellipses)} 个椭圆") for i, ellipse in enumerate(ellipses): ellipse_data = np.array(ellipse).flatten() if len(ellipse_data) < 5: continue center = (int(ellipse_data[0]), int(ellipse_data[1])) axis1 = ellipse_data[2] axis2 = ellipse_data[3] angle = ellipse_data[4] color = (0, 0, 255) # 红色-椭圆 circle_ratio = abs(axis1 - axis2) / max(axis1, axis2) if circle_ratio < 0.1: color = (0, 255, 0) # 绿色-正圆 print(f"椭圆 {i+1}: 中心({center[0]},{center[1]}), 轴长({axis1:.2f},{axis2:.2f}), " f"角度={angle:.2f}°, 圆度={circle_ratio:.4f}") cv.ellipse(frame, center, (int(axis1), int(axis2)), angle, 0, 360, color, 2) cv.circle(frame, center, 3, (255, 0, 0), -1) center_queue.append(center) ellipse_detected = True average_center = None if len(center_queue) > 5: mean_center = np.mean(center_queue, axis=0) std_deviation = np.std(center_queue, axis=0) print(f"\n圆心队列: 数量={len(center_queue)}, 平均值=({mean_center[0]:.2f},{mean_center[1]:.2f}), " f"标准差=({std_deviation[0]:.2f},{std_deviation[1]:.2f})") filtered_centers = [] for center in center_queue: distance = np.linalg.norm(np.array(center) - mean_center) if distance <= 2 * np.linalg.norm(std_deviation): filtered_centers.append(center) if filtered_centers: mean_centers = np.mean(filtered_centers, axis=0) average_center = (int(mean_centers[0]), int(mean_centers[1])) cv.circle(frame, average_center, 6, (255, 0, 0), 2) # 计算偏差 deviation_x = average_center[0] - center_x deviation_y = average_center[1] - center_y print(f"圆心偏差: dx={deviation_x}, dy={deviation_y}") # 绘制偏差线 cv.arrowedLine(frame, (center_x, center_y), average_center, (0, 255, 255), 2) deviation_text = f"dx: {deviation_x:.1f}, dy: {deviation_y:.1f}" cv.putText(frame, deviation_text, (10, 30), cv.FONT_HERSHEY_SIMPLEX, 0.7, (0, 255, 255), 2) # 模糊PID控制 if ellipse_detected: current_h, current_v = servo.get_position() print(f"当前舵机位置: H={current_h}, V={current_v}") # 调用模糊PID控制器 u_x = fuzzy_pid_x.FuzzyPIDcontroller(0, deviation_x) u_y = fuzzy_pid_y.FuzzyPIDcontroller(0, deviation_y) print(f"模糊PID输出: u_x={u_x:.2f}, u_y={u_y:.2f}") # 更新舵机位置 new_h = current_h - u_x * 0.30 new_v = current_v - u_y * 0.3401 print(f"计算新位置: H={new_h:.2f}, V={new_v:.2f}") # 设置舵机位置 servo.set_position(new_h, new_v, move_time=500) pid_text = f"Fuzzy PID: ux={u_x:.1f}, uy={u_y:.1f}" cv.putText(frame, pid_text, (10, 60), cv.FONT_HERSHEY_SIMPLEX, 0.7, (0, 255, 0), 2) # 检查是否达到目标位置 if abs(deviation_x) < 10 and abs(deviation_y) < 10: target_text = "Target Reached!" cv.putText(frame, target_text, (image_width // 2 - 100, 30), cv.FONT_HERSHEY_SIMPLEX, 0.9, (0, 255, 0), 2) # 计算并显示帧率 elapsed_time = time.time() - start_time if elapsed_time > 1: fps = frame_count / elapsed_time frame_count = 0 start_time = time.time() print(f"当前帧率: {fps:.1f} FPS") if 'fps' in locals(): fps_text = f"FPS: {fps:.1f}" cv.putText(frame, fps_text, (10, image_height - 10), cv.FONT_HERSHEY_SIMPLEX, 0.6, (0, 255, 255), 1) # 显示图像 cv.imshow("Detected Ellipses", frame) # 退出处理 if cv.waitKey(1) & 0xFF == ord('q'): board.bus_servo_set_position(1, [[1, 500], [2, 500]]) print("退出程序: 舵机归中") break # 释放资源 cap.release() cv.destroyAllWindows() print("程序结束") if __name__ == '__main__': main()怎么减小曝光值
08-01
- 优化建议：结合其他参数如增益（`cv2.CAP_PROP_GAIN`）和分辨率（`cv2.CAP_PROP_FRAME_WIDTH`）进行综合调整，以提升图像质量[^2]。 #### 常见问题解决 - **曝光值无效？**：运行`v4l2-ctl --all`（Linux）或...
sysinternals程序和部分程序源码
06-20
sysinternals程序和部分程序源码 Sysinternals_Source/ AccessEnumSource filemon434 NtfsInfoSource TokenmonSource AdRestoreSource fmifs procfeatures TVcache AutologonSource FundeleteSource regmon435 VcMonSource CacheSetSource JunctionSource SDeleteSource VxDMonSource Ctrl2CapSource NetstatpSource SecDemoSource DiskExt NewSidSource ShareEnumSource SysinternalsSuite/ accesschk.exe Diskmnt.hlp PHYSMEM.EXE Reghide.exe AccessEnum.exe Diskmon.exe pipelist.exe regjump.exe AdExplorer.chm DISKMON.HLP PORTMON.CNT Regmon.exe ADExplorer.exe DiskView.exe portmon.exe REGMON.HLP ADInsight.chm DMON.SYS PORTMON.HLP RootkitRevealer.chm ADInsight.exe du.exe procexp.chm RootkitRevealer.exe adrestore.exe efsdump.exe procexp.exe sdelete.exe Autologon.exe Eula.txt ProcFeatures.exe ShareEnum.exe autoruns.chm Filemon.exe procmon.chm ShellRunas.exe autoruns.exe FILEMON.HLP Procmon.exe sigcheck.exe autorunsc.exe handle.exe psexec.exe streams.exe Bginfo.exe hex2dec.exe psfile.exe strings.exe Cacheset.exe junction.exe psgetsid.exe sync.exe Clockres.exe ldmdump.exe Psinfo.exe tcpvcon.exe Contig.exe Listdlls.exe pskill.exe tcpview.chm Coreinfo.exe livekd.exe pslist.exe Tcpview.exe ctrl2cap.amd.sys LoadOrd.exe psloggedon.exe TCPVIEW.HLP ctrl2cap.exe logonsessions.exe psloglist.exe Volumeid.exe ctrl2cap.nt4.sys movefile.exe pspasswd.exe whois.exe ctrl2cap.nt5.sys newsid.exe psservice.exe Winobj.exe dbgview.chm ntfsinfo.exe psshutdown.exe WINOBJ.HLP Dbgview.exe pagedfrg.exe pssuspend.exe ZoomIt.exe Desktops.exe pagedfrg.hlp Pstools.chm diskext.exe pdh.dll psversion.txt Diskmnt.exe pendmoves.exe RegDelNull.exe
win10将Capslock变为Ctrl
summer的博客
12-02 3956
之前有使用过 Ctrl2Cap 将Capslock变为Ctrl 一直都很方便使用 管理员权限(主意一定要管理员权限哦)打开命令行运行ctrl2cap.exe会提示使用方法 usage: ctrl2cap.exe [/install | /uninstall] 安装 ctrl2cap.exe /install 卸载 ctrl2cap.exe /uninstall 但是后来在9102年某一次装机win...
微软出品的实用小工具
05-11 4511
分享一些微软出品的实用小工具，希望对大家有所帮助。 包括Sysinternals Suite程序集合、Microsoft安全扫描程序、setspn、Log Parser、Microsoft 网络监视器、数据恢复神器 Windows File Recovery、PowerToys等
获取按键信息
一介码农，热爱金融。
11-19 1922
Windows程序都是基于消息设计的。当你按下一个按键，或者点击一下鼠标，都会想操作系统的内核发送一个Irp消息。你只要捕获这个消息，将这个消息复制下，然后将消息放行。将复制的消息传送给你的应用程序。这就是获取键盘按键消息的实现方法。         很多黑客软件获取别人计算的密码，大多也用这种方法。建议广大用户在使用计算机进行输入游戏账号密码或者银行账号密码等，多用软键盘。防止别人通过这种
评论
成就一亿技术人!
拼手气红包6.0元
还能输入1000个字符
 
红包 添加红包
表情包 插入表情
表情包 代码片
 条评论被折叠 查看
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值