Space Invaders Detector

This application detects Space Invader patterns in radar samples. By using pattern matching with customizable similarity thresholds, it can identify potential invaders even with noise and partial visibility.

Features

• Detect multiple known Space Invader patterns in radar samples
• Multiple detection algorithms with pluggable architecture
• Adjust similarity threshold for detection sensitivity
• Handle partially visible invaders at radar edges
• Configurable via command-line options
• Filter detection by invader types
• Extensible design for adding new invader patterns and algorithms

Requirements

• Ruby 3.4.2 or higher
• Bundler

Installation

1. Clone the repository: bash git clone https://github.com/megarami/space-invaders cd space_invaders

2. Install dependencies: bash bundle install

Usage

Run the detector with a radar sample file:

bash ruby bin/run_detector.rb radar_1.txt

Command Line Options

Usage: run_detector.rb [options] RADAR_FILE -s, --similarity FLOAT Minimum similarity threshold (default: 0.7) -i, --invaders LIST Invader types to detect: large, small, or all (default: large,small) -a, --algorithm ALGORITHM Detection algorithm to use: naive, cross, block (default: naive) --[no-]visualization Enable/disable visualization (default: true) -f, --format FORMAT Output format: text or ascii (default: text) -v, --visibility FLOAT Minimum visibility threshold (default: 0.5) -d, --duplicate-threshold FLOAT Duplicate detection threshold (default: 0.1) -h, --help Show this help message

Examples

Detect all invaders with default settings: bash ruby bin/run_detector.rb samples/radar1.txt

Detect with a lower similarity threshold (more sensitive): bash ruby bin/run_detector.rb -s 0.6 samples/radar_1.txt

Detect only small invaders: bash ruby bin/run_detector.rb -i small samples/radar_1.txt

Use a specific detection algorithm: bash ruby bin/run_detector.rb -a cross samples/radar_1.txt

Disable visualization output: bash ruby bin/run_detector.rb --no-visualization samples/radar_1.txt

Known Invader Patterns

Large Invader

--o-----o-- ---o---o--- --ooooooo-- -oo-ooo-oo- ooooooooooo o-ooooooo-o o-o-----o-o ---oo-oo---

Small Invader

---oo--- --oooo-- -oooooo- oo-oo-oo oooooooo --o--o-- -o-oo-o- o-o--o-o

How It Works

1. The application loads the radar sample and known invader patterns
2. For each invader, it scans the radar by sliding a window across all possible positions
3. At each position, it calculates the similarity between the invader pattern and the radar data
4. If the similarity exceeds the threshold and enough of the pattern is visible, it's considered a match
5. Duplicate matches (same invader in very close positions) are filtered out
6. Results are displayed, showing each match with its position and similarity score

Detection Algorithms

The application supports multiple detection algorithms:

1. Naive Detection Algorithm (naive) - A straightforward sliding window approach that compares each pattern position.
2. Cross Correlation Algorithm (cross) - Uses cross-correlation techniques for more efficient detection.
3. Block Scanning Algorithm (block) - Optimized algorithm that scans in blocks for faster processing.

Adding New Invader Patterns

To add a new invader pattern, simply create a new file in the patterns folder with your pattern.

For example, to add a medium invader pattern, create patterns/medium_invader.txt:

```

Your pattern here using o for invader pixels

and - or spaces for empty pixels

```

Adding New Detection Algorithms

To add a new detection algorithm:

1. Create a new algorithm class in lib/space_invaders/algorithms/ (e.g., enhanced_algorithm.rb):

```ruby

frozenstringliteral: true

module SpaceInvaders class EnhancedAlgorithm < DetectionAlgorithm def detect(radar, invader, threshold) # Your detection logic here end end end ```

1. Register your algorithm with the registry in your initialization code:

ruby SpaceInvaders::AlgorithmRegistry.register('enhanced', EnhancedAlgorithm)

Running Tests

Run the test suite with:

bash bundle exec rspec

To check code quality:

bash bundle exec rubocop

License

This project is licensed under the MIT License - see the LICENSE file for details.