Street Map Generator

📖 Overview

Street Map Generator is a Python-based automation tool that can generate high-resolution city map mosaics.
It uses Selenium to automatically navigate Baidu Maps at specific coordinates, capture screenshots of urban areas, crop them precisely, and merge them into one complete stitched map.

This tool can be used for:

• City visualization and topographic studies
• Urban planning or UI prototyping
• Background map generation for data visualization
• Simulation environments for robotics or autonomous systems

Click HERE to go to the project.
See SHANGHAI.png as an example. The picture without blackline can serve as an example.

---

🧩 Features

• 🌐 Automated map capture using Selenium and ChromeDriver
• 📸 Precise image cropping with OpenCV to remove browser frames
• 🧱 Automatic image stitching via Pillow
• ⚙️ Configurable coordinate range and resolution
• 🧮 Extensible coordinate base for different cities (e.g., Shanghai, Suzhou)

---

🧠 File Descriptions

File	Description
get_pic.py	Main script — handles map fetching, cropping, and stitching
visit.py	Early version for coordinate-based automated browsing
click_test.py	Mouse coordinate debugging tool for calibration

---

🧰 Dependencies

Install required packages via pip:

pip install selenium opencv-python pillow tqdm pyautogui pywin32

You will also need:

• Google Chrome
• ChromeDriver (version must match your Chrome browser)

---

⚙️ Usage

1️⃣ Set Parameters

In get_pic.py, adjust parameters such as base coordinates and step size:

# Base coordinates for Baidu Map
BASE_X = 13386335    # Suzhou
BASE_Y = 3583355

# Step size in map units
STEP_X = 7000
STEP_Y = 4000

# Capture grid size
for index_x in range(19):      # horizontal range
    for index_y in range(41):  # vertical range
        get_original_pic(index_x, index_y)

2️⃣ Run the Program

python get_pic.py

The script will:

1. Automatically open Chrome and navigate to target coordinates
2. Take screenshots and save them to pics3/
3. Automatically stitch them into one large image, e.g. Suzhou.png

---

🖼️ Output Example

The final resolution depends on the number of rows and columns as well as image dimensions.
Example: with 19 × 41 tiles of 1236×706 pixels → about 23,484 × 28,946 px stitched image.

---

⚠️ Notes

• Excessive requests may trigger anti-bot mechanisms — use random sleep() delays.
• The script will open many browser windows; do not close them manually.
• Works best on Windows (uses win32api and ctypes.windll for mouse control).
  Mac/Linux users need to replace mouse-related calls.