README.md

---
title: Farm Segmentation API
emoji: 🏞️
colorFrom: green
colorTo: blue
sdk: gradio
sdk_version: 4.28.3
app_file: app.py
pinned: false
license: apache-2.0
short_description: AI-powered agricultural image segmentation and land analysis
---

# 🏞️ Farm Segmentation API

Advanced agricultural image segmentation using SegFormer models for precise field and crop analysis.

## 🎯 Capabilities
- **Semantic Segmentation**: Pixel-level classification of agricultural scenes
- **Agricultural Categories**: Soil, vegetation, water, buildings, equipment
- **Composition Analysis**: Percentage breakdown of field components
- **Multi-Resolution**: Support for different accuracy/speed tradeoffs

## 🤖 Models
- **SegFormer B0**: Fastest processing, basic accuracy
- **SegFormer B1**: Balanced performance (recommended)
- **SegFormer B2**: Highest accuracy, slower processing

## 📡 API Usage

### Python
```python
import requests
import base64

def segment_farm_image(image_path, model="segformer_b1"):
    with open(image_path, "rb") as f:
        image_b64 = base64.b64encode(f.read()).decode()
    
    response = requests.post(
        "https://YOUR-USERNAME-farm-segmentation.hf.space/api/predict",
        json={"data": [image_b64, model]}
    )
    return response.json()

result = segment_farm_image("field_image.jpg")
print(result)
```

## 📊 Response Format
```json
{
  "segments_detected": 8,
  "segments": [
    {
      "class": "grass",
      "agricultural_category": "vegetation",
      "pixel_count": 145632,
      "percentage": 35.2,
      "label_id": 9
    },
    {
      "class": "soil",
      "agricultural_category": "soil", 
      "pixel_count": 98234,
      "percentage": 23.7,
      "label_id": 12
    }
  ],
  "processing_time": 2.1
}
```

## 🌾 Agricultural Categories
- **soil**: Ground, dirt, earth, mud
- **vegetation**: Crops, grass, trees, plants
- **water**: Irrigation channels, ponds, rivers
- **building**: Barns, greenhouses, structures
- **equipment**: Tractors, machinery, tools
- **other**: Roads, sky, uncategorized objects