Installation Guide (Linux)
This guide provides detailed instructions on how to deploy and run MipMapEngine SDK on Linux systems.
đ System Requirementsâ
Recommended Configurationâ
| Component | Requirement |
|---|---|
| Operating System | Ubuntu 22.04 |
| CPU | Intel/AMD x86_64 8 cores or more |
| Memory | 32 GB RAM |
| GPU | NVIDIA GPU (CUDA support) |
| Graphics Driver | NVIDIA Driver >= 525.60.33 |
| CUDA | CUDA 12.8 |
| Disk Space | 1 TB available space |
đ Installation Stepsâ
Step 1: Check Hardware and Driversâ
Check GPU Informationâ
# Check if NVIDIA GPU is installed
lspci | grep -i nvidia
# Check NVIDIA driver version
nvidia-smi
If drivers are not installed or version is too low:
# Ubuntu install NVIDIA driver
sudo apt update
sudo apt install nvidia-driver-***
Step 2: Install Virbox Driverâ
Virbox is the SDK's encryption tool and must be installed.
# Install Virbox user tools
https://lm.virbox.com/tools.html
Step 3: Activate Licenseâ
# Use SDK provided license_engine to activate
./license_engine -bind YOUR-LICENSE-KEY
# Query license status
./license_engine -enum license_info.json
cat license_info.json
Step 4: Extract SDKâ
# Extract SDK package
tar -xzvf MipMapEngine_SDK_Linux_v4.0.tar.gz
# View directory structure
cd MipMapEngine_SDK
ls -la
# Should contain:
# - bin/ # Executables and dynamic libraries
# - gdal_data/ # GDAL data files
# - docs/ # User manual
# - examples/ # Example scripts
Step 5: Configure Environment Variablesâ
# Set SDK path (replace with actual path)
export MIPMAP_SDK_PATH=/path/to/MipMapEngine_SDK
# Set library path
export LD_LIBRARY_PATH=$MIPMAP_SDK_PATH/bin:$LD_LIBRARY_PATH
# Recommend adding to ~/.bashrc or ~/.profile
echo 'export MIPMAP_SDK_PATH=/path/to/MipMapEngine_SDK' >> ~/.bashrc
echo 'export LD_LIBRARY_PATH=$MIPMAP_SDK_PATH/bin:$LD_LIBRARY_PATH' >> ~/.bashrc
source ~/.bashrc
Step 6: Check Dependenciesâ
# Check executable dependencies
ldd $MIPMAP_SDK_PATH/lib/reconstruct_full_engine
# If libraries are missing, install system packages
# Ubuntu
apt-get update && apt-get install -y libgl1-mesa-dev libgeos-dev libgomp1 liblcms2-2 libquadmath0 libjbig0 libdeflate0 libopenexr25 libproj22
Step 7: Verify Installationâ
Create test configuration file test.json:
{
"working_dir": "./test_output",
"gdal_folder": "/path/to/MipMapEngine_SDK/gdal_data",
"input_image_type": 1,
"resolution_level": 3,
"coordinate_system": {
"type": 2,
"epsg_code": 4326
},
"image_meta_data": [
{
"id": 1,
"path": "/path/to/test/image1.jpg"
},
{
"id": 2,
"path": "/path/to/test/image2.jpg"
}
]
}
Run test:
# Enter SDK directory
cd $MIPMAP_SDK_PATH/bin
# Run reconstruction
./reconstruct_full_engine -reconstruct_type 0 -task_json test.json
# Check output
ls -la ./test_output/
đ§ Common Issuesâ
1. Missing libmipmap_engine.soâ
Error message:
error while loading shared libraries: libmipmap_engine.so: cannot open shared object file
Solution:
# Ensure library path is correct
export LD_LIBRARY_PATH=$MIPMAP_SDK_PATH/bin:$LD_LIBRARY_PATH
2. CUDA-related Errorsâ
Error message:
CUDA error: no CUDA-capable device is detected
Solution:
# Check GPU drivers
nvidia-smi
# Update drivers
sudo apt update
sudo apt upgrade nvidia-driver-*
4. Library Version Conflictsâ
Some system libraries may conflict with SDK-provided libraries:
# Identify conflicting libraries
ldd -v $MIPMAP_SDK_PATH/bin/reconstruct_full_engine
# Temporarily remove conflicting libraries
mv $MIPMAP_SDK_PATH/libconflict.so $MIPMAP_SDK_PATH/libconflict.so.bak
# Use system libraries
# System will automatically load from /usr/lib
Next Stepsâ
- Check Quick Start to run your first project
- Read API Reference for detailed interfaces
- Refer to Troubleshooting for common issues
Tip: Linux deployment may require adjustments due to distribution and environment differences. It's recommended to verify in a test environment before deploying to production.