Loading repository data…
Loading repository data…
DEVILENMO / repository
基于Yolo的辅助瞄准系统，高度模块化编写，也可以用于光学控制、监控物体追踪等领域，持续更新中，仅供交流学习使用。An auto-aiming system, or aim bot, which is mainly based on Yolo. Can be easily used in different areas such as optical control, monitoring object tracking. I will keep fixing bugs and adding new features.
A transparent discovery signal based on current public GitHub metadata.
This score does not audit code, security, maintainers, documentation quality, or suitability. Verify the repository and its current documentation before adoption.
DeadEye 目标追踪系统是一款高性能的图像辅助目标追踪工具。作为示例，提供了在FPS游戏中的目标追踪以及对应的瞄准效果演示。示例程序通过分析屏幕截图，检测并追踪目标，最后通过模拟操作实现辅助瞄准和自动扳机功能。不仅如此，我还将这个系统应用在光学激光控制和监控探头的目标追踪中，虽然这些领域的示例程序没有提供，但用户可以通过简单修改相关模块，将这个系统应用于不同领域。
DeadEye target tracking system is a high-performance image-assisted target tracking tool. As an example, it provides a demonstration of target tracking and corresponding aiming effects in FPS games. The sample program analyzes screenshots, detects and tracks targets, and finally implements assisted aiming and automatic trigger functions through simulated operations. Additionally, I have applied this system to optical laser control and monitoring target tracking. Although sample programs for these fields are not provided, users can apply this system in different domains by simply modifying the corresponding modules.
DeadEyeCore（deadeye/core/dead_eye_core.py）中数据流为：输入源取帧 → DetectModule.target_detect → 匈牙利匹配与卡尔曼更新 Target 列表 → 依次调用各 ExecutionModule.update_targets(frame, target_list)。*_module：deadeye/input_module、deadeye/detection_module、deadeye/execution_module（原 input / detection / aim / output 已调整；VideoWriterOutput 已并入 execution_module）。ExecutionModule，在检测与追踪完成后对当前帧与 target_list 调用 update_targets()；核心类 DeadEyeCore 使用 execution_modules 列表按顺序执行。DeadEyeAutoAimingModule 或 VideoWriterOutput 等。VideoFileCamera 改为选择单个视频文件（video_path），不再扫描整个文件夹。main.py 扫描 *_module 包与 ConfigurableMixin 动态生成参数表单。DeadEyeCore (deadeye/core/dead_eye_core.py), the flow is: grab a frame from the input camera → DetectModule.target_detect → Hungarian matching + Kalman updates on Target → each ExecutionModule.update_targets(frame, target_list) in order.*_module: deadeye/input_module, deadeye/detection_module, deadeye/execution_module (replacing the old input / detection / aim / output layout; VideoWriterOutput now lives under execution_module).ExecutionModule base class and implement update_targets(); DeadEyeCore runs a list execution_modules in order.DeadEyeAutoAimingModule or VideoWriterOutput.VideoFileCamera uses a single file (video_path), not a directory of videos.main.py discovers classes under *_module packages and builds ConfigurableMixin forms for their settings..pt 权重文件的直接检测或使用 TensorRT 部署后的 .trt / .engine 模型进行精确的目标检测。DeadEyeAutoAimingModule、VideoWriterOutput 等继承 ExecutionModule，在追踪更新后对帧与目标列表执行瞄准、写带框视频等动作；辅助瞄准使用 PID 控制算法。input_module、detection_module、execution_module 中按基类扩展；鼠标等行为由执行模块内 MouseControlModule 等实现，按需自行接入硬件或驱动。.pt weight files or precise target detection using deployed TensorRT models with .trt / .engine formats.DeadEyeAutoAimingModule and VideoWriterOutput inherit ExecutionModule and run after tracking updates (aim assist, annotated video export, etc.); aim assist uses a PID controller.input_module, detection_module, and execution_module via their base classes; mouse or other actuation is wired inside execution modules (e.g. MouseControlModule patterns) for your own hardware.deadeye/input_module作为示例实现了快速的截图相机 SimpleScreenShotCamera 以及单文件视频源 VideoFileCamera（选择一个视频路径）。用户可继承 BaseCamera 实现 get_image()，通过 main.py 的 UI 或自行组装 DeadEyeCore 使用。
dxcam 或 mss 实现高速截图功能。BaseCamera 类并且实现 get_image() 函数返回恰当的图像就可以自定义你的相机了。deadeye/detection_module作为示例实现了基于 Yolov8 的 YoloDetector 类（支持 ConfigurableMixin，模型路径等在模块参数中配置）。用户可实现自己的检测类，保证输出格式与现有流水线一致即可。
.pt 模型进行直接检测，也支持使用 .trt / .engine 模型的基于 TensorRT 的部署方式进行高效检测。deadeye/execution_module示例包括 DeadEyeAutoAimingModule（PID 辅助瞄准）与 VideoWriterOutput（画框并写出视频）。均继承 ExecutionModule，实现 update_targets(image, target_list)。
ExecutionModule 的类，放入 execution_module 包后由 UI 自动发现（需满足扫描规则）。DeadEyeAutoAimingModule 类会在 __init__ 时实例化这个类并用这个类来控制鼠标。用户需要自己实现这个类里控制鼠标两个函数 click_left_button 以及 move_mouse ，分别对应点击鼠标左键以及移动鼠标。deadeye/input_moduleExample implementations include SimpleScreenShotCamera for fast screen capture and VideoFileCamera for a single video file path. Inherit BaseCamera, implement get_image(), and wire it through the main.py UI or by constructing DeadEyeCore manually.
dxcam or mss to achieve high-speed screenshot functionality.BaseCamera class and implement the get_image() function to return the appropriate image.deadeye/detection_moduleYoloDetector is based on Yolov8 and supports ConfigurableMixin (model path and other options in module settings). You may provide your own detector class as long as its output matches the pipeline contract.
.pt models for direct detection and also supports efficient detection using .trt / .engine models through TensorRT deployment.deadeye/execution_moduleExamples include DeadEyeAutoAimingModule (PID aim assist) and VideoWriterOutput (draw boxes and write video). All inherit ExecutionModule and implement update_targets(image, target_list).
execution_module so the UI discovery can pick them up (subject to the package scan rules).DeadEyeAutoAimingModule class will instantiate this class during __init__ and use this class to control the mouse. Users need to implement the two mouse control functions in this class, click_left_button and move_mouse, which correspond to clicking the left mouse button and moving the mouse respectively.main.py 启动图形界面：在「基本设置」中切换语言；在「启动配置」中选择检测 / 输入 / 执行模块并填写参数（如 YOLO 模型路径、视频文件路径等）；在「运行控制」查看状态与 FPS，使用 开始运行 / 暂停 / 继续 控制流水线。仍可使用快捷键暂停/继续、退出，以及运行时切换辅助瞄准、自动扳机等选项。P : 暂停/继续程序O : 完全结束程序鼠标左键 : 按下时开启自动瞄准main.py to open the UI. Use Basic settings (language), Pipeline configuration (detector / input / execution modules and fields such as YOLO model path and video file path), then Run control (status, FPS, Start / Pause / Resume). Hotkeys still pause/resume/exit; runtime toggles include auto-aim and auto-trigger when available.P : Pause/Resume the programO : Completely exit the programLeft Mouse Button : Enable auto-aiming while pressed以下动图展示了此项目可提供的辅助瞄准效果：
The following animation demonstrates the aiming assist effect provided by this project:
ExecutionModule、main.py UI 布局与多语言文案等），细节见上文 「近期更新（2026）」；与旧教程或截图不一致时以当前仓库代码为准。