Ahnuf/Military_Aircraft_Detection_Classification_Image_Dataset

1---2license: apache-2.03task_categories:4  - object-detection5  - image-classification6tags:7  - military8  - aircraft9  - aerospace10  - yolo11  - defense12  - commercial-aircraft13  - birds14  - drones15size_categories:16  - 10K-100K17---18 19# Military Aircraft Detection & Classification Dataset  20### 88 Classes with Advanced Background Suppression21 22## Overview23This dataset is a professionally curated resource for training high-performance **object detection** and **image classification** models such as **YOLOv11**.  24It contains **88 distinct military aircraft classes** and is explicitly designed for **real-world deployment**, where false positives from civilian aircraft, birds, and small drones are common.25 26To address this, the dataset incorporates a structured **background suppression strategy**, teaching models not only what *to detect*, but also what *to ignore*.27 28---29 30## Dataset Summary31 32- **Total Images**: 26,668  33- **Military Aircraft Classes**: 87  34- **Image Resolution**: 640 × 640 (uniform)  35- **Annotation Format**: YOLO (`.txt`) with normalized coordinates  36- **Primary Use**: Military aircraft detection and classification  37 38---39 40## Dataset Split & Background Statistics41 42| Split | Total Images | Background Images | Background % |43|------|-------------|------------------|--------------|44| Train | 21,342 | 2,508 | 11.75% |45| Validation | 2,641 | 295 | 11.17% |46| Test | 2,645 | 284 | 10.74% |47| **Total** | **26,668** | **3,127** | **~11.7%** |48 49The dataset maintains a **stratified split** of approximately **80% Train / 10% Validation / 10% Test** across all 87 classes.50 51---52 53## Advanced Background Suppression Strategy54 55To significantly reduce false detections, the dataset includes **3,127 background-only images** with **empty annotations**.  56These images are intentionally selected to represent common real-world confounders in aerial imagery.57 58### Background Categories59 601. **Empty Skies, Clouds & Commercial Aircraft**  61   Negative samples containing:62   - Clear or cloudy skies with no aircraft  63   - **Commercial passenger and cargo aircraft**  64   This trains the model to distinguish civilian airliners from military platforms.65 662. **Bird Backgrounds (≈1.5%)**  67   High-resolution bird imagery to prevent *bird-as-aircraft* false positives, particularly at long range or low resolution.68 693. **Commercial Drone Backgrounds (≈1.5%)**  70   Civilian and hobbyist UAVs (quadcopters and small drones), enabling the model to differentiate between commercial drones and military-grade UAVs.71 72All background images use **empty `.txt` label files (0 bytes)** and contain **no bounding boxes**.73 74---75 76## Annotation Format77 78Each image is paired with a corresponding `.txt` file in YOLO format.79 80### Positive Sample Example81 82`su57_01.txt`8368 0.475000 0.496875 0.415625 0.85937584 85**Field Description**86- `68` → Class ID (Su-57)  87- `0.475000` → X-center (47.5% of image width)  88- `0.496875` → Y-center (49.69% of image height)  89- `0.415625` → Bounding box width  90- `0.859375` → Bounding box height  91 92### Background (Negative) Samples93 94Background label files are **intentionally empty**:95- `sky_bg_01.txt`96- `commercial_aircraft_bg_01.txt`97- `birds_v1_01.txt`98- `drones_v1_01.txt`99 100---101 102## Final Class ID Table (87 Classes)103 104| ID | Class | ID | Class | ID | Class | ID | Class |105|----|-------|----|-------|----|-------|----|-------|106| 0 | A10 | 22 | CL415 | 44 | JF17 | 66 | Su34 |107| 1 | A400M | 23 | E2 | 45 | JH7 | 67 | Su47 |108| 2 | AG600 | 24 | E7 | 46 | KAAN | 68 | Su57 |109| 3 | AH64 | 25 | EF2000 | 47 | KC135 | 69 | TB001 |110| 4 | AKINCI | 26 | EMB314 | 48 | KF21 | 70 | TB2 |111| 5 | AV8B | 27 | F117 | 49 | KJ600 | 71 | Tejas |112| 6 | An124 | 28 | F14 | 50 | Ka27 | 72 | Tornado |113| 7 | An22 | 29 | F15 | 51 | Ka52 | 73 | Tu160 |114| 8 | An225 | 30 | F16 | 52 | MQ9 | 74 | Tu22M |115| 9 | An72 | 31 | F18 | 53 | Mi24 | 75 | Tu95 |116| 10 | B1 | 32 | F2 | 54 | Mi26 | 76 | U2 |117| 11 | B2 | 33 | F22 | 55 | Mi28 | 77 | UH60 |118| 12 | B52 | 34 | F35 | 56 | Mi8 | 78 | US2 |119| 13 | Be200 | 35 | F4 | 57 | Mig29 | 79 | V22 |120| 14 | C1 | 36 | FCK1 | 58 | Mig31 | 80 | Vulcan |121| 15 | C130 | 37 | H6 | 59 | Mirage2000 | 81 | WZ7 |122| 16 | C17 | 38 | Il76 | 60 | P3 | 82 | X32 |123| 17 | C2 | 39 | J10 | 61 | RQ4 | 83 | XB70 |124| 18 | C390 | 40 | J20 | 62 | Rafale | 84 | Y20 |125| 19 | C5 | 41 | J35 | 63 | SR71 | 85 | YF23 |126| 20 | CH47 | 42 | J36 | 64 | Su24 | 86 | Z10 |127| 21 | CH53 | 43 | JAS39 | 65 | Su25 | 87 | Z19 |128 129---130 131## Intended Use Cases132- Military aircraft detection and classification  133- Civilian vs military aircraft discrimination  134- UAV and drone differentiation  135- Long-range aerial surveillance research  136- False-positive suppression benchmarking for YOLO models  137