Skip to content

README.md

Latest commit

 

History

History
42 lines (27 loc) · 1.58 KB

README.md

File metadata and controls

42 lines (27 loc) · 1.58 KB
Raw

DriverAssist

Lane and Stop Sign Detection

This project provides a C++ implementation for detecting lanes and stop signs in a video using OpenCV and a custom channel implementation for inter-thread communication.

Features

  • Lane Detection: Detects and highlights lanes on the road using the Hough Line Transform.
  • Stop Sign Detection: Detects stop signs using a pre-trained Haar Cascade Classifier.
  • Multithreading: Uses custom Channel and BufferedChannel classes for inter-thread communication to efficiently process video frames in parallel.

Code Overview

channel.h

Defines Channel and BufferedChannel classes for thread-safe communication.

  • Channel: A simple unbuffered channel for single-item communication.
  • BufferedChannel: A buffered channel that can hold multiple items.

main.cpp

  • Loads video frames.
  • Processes frames for lane and stop sign detection.
  • Uses multithreading for efficient processing.

Functions

  • frameLoader: Loads frames from the video and sends them through a BufferedChannel.
  • processStopSigns: Detects stop signs in a given frame.
  • processLanes: Detects lanes in a given frame and calculates the offset from the lane center.

Configuration

Several preprocessor directives control the behavior of the application:

  • DEBUG: Enable to draw detected lanes on the frames.
  • OUTPUT_VIDEO: Enable to save the processed video.
  • OUTPUT_FRAMES: Enable to save individual frames.
  • RESIZE: Enable to resize frames for faster processing.
  • SHOW_ROI: Enable to show the region of interest.