Skip to content

README.md

Latest commit

 

History

History
70 lines (60 loc) · 2.38 KB

README.md

File metadata and controls

70 lines (60 loc) · 2.38 KB
Raw

Dataset Format Guide

This document describes the stereo image dataset format used by the Visual Odometry and SLAM system. The main dataset is stored in the stereo_rgb directory at the project root (not in this data directory).

Stereo RGB Dataset Structure

stereo_rgb/
├── image_02/                # Left camera images
│   ├── data/                # Image files
│   │   ├── 0000000000.png
│   │   ├── 0000000001.png
│   │   └── ...
│   └── timestamps.txt       # Timestamps for each image
├── image_03/                # Right camera images
│   ├── data/                # Image files
│   │   ├── 0000000000.png
│   │   ├── 0000000001.png
│   │   └── ...
│   └── timestamps.txt       # Timestamps for each image
└── ground_truth.txt         # Ground truth trajectory data

Camera Calibration Parameters

The camera calibration parameters are now stored in the config.yaml file in the project root directory. This file contains the following camera parameters in OpenCV YAML format:

%YAML:1.0
---
camera:
    fx: 718.856
    fy: 718.856
    cx: 624.0
    cy: 192.0
    P0: !!opencv-matrix
       rows: 3
       cols: 4
       dt: d
       data: [ 718.8560, 0.0000, 607.1928, 0.0000,
               0.0000, 718.8560, 185.2157, 0.0000,
               0.0000, 0.0000, 1.0000, 0.0000 ]
    P1: !!opencv-matrix
       rows: 3
       cols: 4
       dt: d
       data: [ 718.8560, 0.0000, 607.1928, -386.1448,
               0.0000, 718.8560, 185.2157, 0.0000,
               0.0000, 0.0000, 1.0000, 0.0000 ]

aliked:
    model_name: "aliked-n32"
    device: "cuda"
    top_k: -1
    scores_thres: 0.2
    n_limit: 3000

stereo_onnx:
  model_path: "/home/saqib/visual_slam/visual_odometry/StereoModel.onnx"  # Path to ONNX stereo model
  input_width: 1248                 # Model input width (must match ONNX model)
  input_height: 384                 # Model input height (must match ONNX model)
  max_disp: 192                     # Maximum disparity value
  baseline: 0.54                    # Stereo baseline in meters (KITTI default)
  use_gpu: 1                        # Use GPU acceleration (1=true, 0=false)
  gpu_device_id: 0                  # GPU device ID
  num_threads: 4

The config file also contains additional parameters for feature detection, tracking, and stereo depth estimation.