FGVC-Aircraft contains 10,200 images of aircraft, with 100 images for each of 102 different aircraft model variants, most of which are airplanes. The (main) aircraft in each image is annotated with a tight bounding box and a hierarchical airplane model label. Aircraft models are organized in a four-levels hierarchy. The four levels, from finer to coarser, are:

Model, e.g. Boeing 737-76J. Since certain models are nearly visually indistinguishable, this level is not used in the evaluation. Variant, e.g. Boeing 737-700. A variant collapses all the models that are visually indistinguishable into one class. The dataset comprises 102 different variants. Family, e.g. Boeing 737. The dataset comprises 70 different families. Manufacturer, e.g. Boeing. The dataset comprises 41 different manufacturers. The data is divided into three equally-sized training, validation and test subsets.

This dataset preprocessed version of: user : a2015003713 dataset: militaryaircraftdetectiondataset

I have done this version for my use and wanted to share if some people wants to use this dataset to training in YOLO. You can contact with me for any question from my website.

Authors introduce the Military Aircraft Detection Dataset, a comprehensive dataset designed for object detection of military aircraft. This dataset features bounding boxes in PASCAL VOC format (xmin, ymin, xmax, ymax) and includes images of 43 distinct aircraft types, such as A-10, F-35, Su-57, and more. The dataset, comprising 12,008 images in total, was sourced from Wikimedia Commons and Google Image Search, making it a valuable resource for training and evaluating object detection models for military aircraft recognition task.

Aircraft Crack

## Overview

Aircraft Crack is a dataset for object detection tasks - it contains Crack annotations for 835 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Aircraft Exterior

## Overview

Aircraft Exterior is a dataset for object detection tasks - it contains Tag Exterior annotations for 4,835 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

The number of aircraft in the United States increased in 2021, estimates holding that the general aviation fleet was ******* aircraft. On the other hand, the for-hire carrier fleet decreased to ***** aircraft. However, it is predicted that the number of for-hire carrier aircraft will slightly increase in 2022, reaching *****. General aviation versus for-hire carriersThe airline industry in the United States is generally divided into two categories: for-hire carriers and general aviation. For-hire carries includes commercial services where an operator agrees to transport passengers, cargo or mail for a fee. General aviation includes basically all others forms of air travel, such as business/personal travel on privately owned aircraft, recreational flying, and other various tasks such as search & rescue, surveying, and photography (among others). Pilots for for-hire carriers are required obtain an ‘airline transport’ license – the highest level of pilot qualification - and are therefore outnumbered by pilots who are qualified only for general aviation. Aviation in the United StatesThe U.S. has one of the largest aviation market of any single country in the world. U.S. airlines transport more passengers than airlines from any other country, and the U.S. is home to around half of the top ten airlines in the world in terms of revenue. This dominance extends into the general aviation sector, with North America having a significantly larger fleet of aircraft for private air travel than any other region. However, when understood at a regional level, the Asia-Pacific region currently maintains a slightly larger commercial aircraft fleet, which is expected to become significantly larger than the North American fleet over the next 20 years.

The National Aeronautics and Space Administration (NASA) Aircraft Scanners data set contains digital imagery acquired from several multispectral scanners, including Daedalus thematic mapper simulator scanners and the thermal infrared multispectral scanner. Data are collected from selected areas over the conterminous United States, Alaska, and Hawaii by NASA ER-2 and NASA C-130B aircraft, operating from the NASA Ames Research Center in Moffett Field, California, and by NASA Learjet aircraft, operating from Stennis Space Center in Bay St. Louis, Mississippi. Limited international acquisitions also are available. In cooperation with the Jet Propulsion Laboratory and Daedalus Enterprises,Inc., NASA developed several multispectral sensors. The data acquired from these sensors supports NASA's Airborne Science and Applications Program and have been identified as precursors to the instruments scheduled to fly on Earth Observing System platforms. THEMATIC MAPPER SIMULATOR The Thematic Mapper Simulator (TMS) sensor is a line scanning device designed for a variety of Earth science applications. Flown aboard NASA ER-2 aircraft, the TMS sensor has a nominal Instantaneous Field of View of 1.25 milliradians with a ground resolution of 81 feet (25 meters) at 65,000 feet. The TMS sensor scans at a rate of 12.5 scans per second with 716 pixels per scan line. Swath width is 8.3 nautical miles (15.4 kilometers) at 65,000 feet while the scanner's Field of View is 42.5 degrees. NS-001 MULTISPECTRAL SCANNER The NS-001multispectral scanner is a line scanning device designed to simulate Landsat thematic mapper (TM) sensor performance, including a near infrared/short-wave infrared band used in applications similar to those of the TM sensor (e.g., Earth resources mapping, vegetation/land cover mapping, geologic studies). Flown aboard NASA C-130B aircraft, the NS-001 sensor has a nominal Instantaneous Field of View of 2.5 milliradians with a ground resolution of 25 feet (7.6 meters) at 10,000 feet. The sensor has a variable scan rate (10 to 100 scans per second) with 699 pixels per scan line, but the available motor drive supply restricts the maximum stable scan speed to approximately 85 revolutions per second. A scan rate of 100 revolutions per second is possible, but not probable, for short scan lines; therefore, a combination of factors, including aircraft flight requirements and maximum scan speed, prevent scanner operation below 1,500 feet. Swath width is 3.9 nautical miles (7.26 kilometers) at 10,000 feet, and the total scan angle or field of regard for the sensor is 100 degrees, plus or minus 15 degrees for roll compensation. THERMAL INFRARED MULTISPECTRAL SCANNER The Thermal Infrared Multispectral Scanner (TIMS) sensor is a line scanning device originally designed for geologic applications. Flown aboard NASA C-130B, NASA ER-2, and NASA Learjet aircraft, the TIMS sensor has a nominal Instantaneous Field of View of 2.5 milliradians with a ground resolution of 25 feet (7.6 meters) at 10,000 feet. The sensor has a selectable scan rate (7.3, 8.7, 12, or 25 scans per second) with 698 pixels per scan line. Swath width is 2.6 nautical miles (4.8 kilometers) at 10,000 feet while the scanner's Field of View is 76.56 degrees.

Aircraft collision avoidance systems rely on sensor information to detect and track intruding aircraft so that they may issue proper collision avoidance advisories. While typical surveillance sensors for manned aircraft include transponders and onboard radar, autonomous aircraft will require additional sensors both for redundancy and to replace the visual acquisition typically performed by the pilot. As a result, the community has proposed detecting other aircraft using vision-based sensors such as cameras. These sensors require the development of techniques to process images of the environment to detect intruding aircraft. To boost this development, this artifact provides a dataset of 72,000 labeled images of intruder aircraft with various lighting conditions, weather conditions, relative geometries, and geographic locations. For more information on the structure of this dataset as well as benchmark models and a full simulator, see https://github.com/sisl/VisionBasedAircraftDAA.

Click to add a brief description of the dataset (Markdown and LaTeX enabled).

Provide:

a high-level explanation of the dataset characteristics explain motivations and summary of its content potential use cases of the dataset

The world's aircraft fleet is expected to increase from ****** to ****** aircraft between 2023 and 2043. While the more established markets of Eurasia and North America were predicted to increase by around ** and ** percent respectively, the Chinese fleet is expected to increase by about *** percent to ***** aircraft in 2043.

The ACES Aircraft and Mechanical Data consist of aircraft (e.g. pitch, roll, yaw) and mechanical (e.g. aircraft engine speed, tail commands, fuel levels) data recorded by the Altus II Unmanned Aerial Vehicle (Altus II UAV) system during the Altus Cumulus Electrification Study (ACES) based at the Naval Air Facility Key West in Florida. ACES aimed to provide extensive observations of the cloud electrification process and its effects by using the Altus II UAV to collect cloud top observations of thunderstorms. The campaign also worked to validate satellite lightning measurements. The Altus II aircraft and mechanical data files are available from July 10 through August 30, 2002 in MATLAB data format (.mat).

This dataset provides flight track and aircraft navigation data from the NASA Atmospheric Tomography Mission (ATom). Flight track information is available for the four ATom campaigns: ATom-1, ATom-2, ATom-3, and ATom-4. Each ATom campaign consists of multiple individual flights and flight navigational information is recorded in 10-second intervals. Data available for each flight includes research flight number, date, and start and stop time of each 10-second interval. In addition, latitude, longitude, altitude, pressure and temperature is included at each 10-second interval. NASA's ATom campaign deploys an extensive gas and aerosol payload on the NASA DC-8 aircraft for systematic, global-scale sampling of the atmosphere, profiling continuously from 0.2 to 12 km altitude. Flights occurred in each of 4 seasons from 2016 to 2018. During each campaign, flights originate from the Armstrong Flight Research Center in Palmdale, California, fly north to the western Arctic, south to the South Pacific, east to the Atlantic, north to Greenland, and return to California across central North America. ATom establishes a single, contiguous, global-scale dataset. One intended use of this flight track data is to facilitate to mapping model results from global models onto the precise ATom flight tracks for comparison.

Graph and download economic data for Producer Price Index by Industry: Aircraft Manufacturing: Civilian Aircraft (PCU3364113364113) from Dec 1985 to May 2025 about aircraft, civilian, manufacturing, PPI, industry, inflation, price index, indexes, price, and USA.

Commercial Aircraft New

## Overview

Commercial Aircraft New is a dataset for object detection tasks - it contains Aircraft annotations for 2,105 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

It is estimated that the U.S. Navy will have a total of 1,002 strike fighters in FY 2023, slightly lower than the 1,009 strike fighters they maintained in FY 2022. These values include active and reserve aircraft from both the Navy and Marines inventories.

A. SUMMARY

This dataset contains the aircraft tail and model information recorded by the Aircraft Parking System at SFO.

The columns in the dataset include details such as tail number, model, airline, status, as well as the Creation and Modification Dates of the aircraft record.

On the 28th day of each month, this dataset is exported from the Aircraft Parking System at SFO.

B. HOW THE DATASET IS CREATED

When airline requests for reservation of aircraft parking stand at SFO, a new aircraft parking reservation is entered to the Aircraft Parking System. If the requested Aircraft is not found in the Aircraft Parking System, then the requested Aircraft information is added to the aircraft table of the Aircraft Parking System.

On the 28th day of each month, this dataset is replaced with all the records retrieved from the aircraft table of the Aircraft Parking System at SFO.

C. UPDATE PROCESS

When airline companies undergo a merger or acquisition, the transfer of aircraft ownership is recorded by marking the aircraft previously owned by the originating airline as inactive and registering the aircraft now owned by the new airline as active in the aircraft table.

On the 28th day of each month, this dataset is replaced with all the records retrieved from the aircraft table of the Aircraft Parking System at SFO.

E. RELATED DATASETS Aircraft Parking Activity Records at SFO Aircraft Parking Location Inventory at SFO

Utilize the Tail Number column of the "Aircraft Parking Activity Records at SFO" dataset to connect with the Tail Number column of the "Aircraft Tail Numbers and Models at SFO" dataset, for referencing the corresponding aircraft record.

The industry experienced a strong decline during the pandemic, with a staggering 70% drop in air travel primarily driven by social distancing measures and a global economic shock. This led to substantial revenue losses for airlines and lowered demand for new aircraft. However, the industry began recovering post-pandemic, aided by relaxed restrictions, economic stimulus packages and lower interest rates, which made borrowing for capital more affordable. Ultimately, the industry reached double-digit growth in 2023 but slowed down to 4.6% growth in 2024 due to mounting logistical and economic challenges. At an estimated $321.1 billion in 2024 revenues, the industry remains below the pre-pandemic levels despite accelerated growth seen in the second half of the current period, which can be attributed to logistical challenges that major players are experiencing despite strong demand for new aircraft. For instance, the grounding of Boeing 737-9 MAX aircraft and issues at engine supplier Pratt & Whitney have created high backlogs for manufacturers like Boeing, Airbus and Embraer. As a result of the pandemic and the inability to ramp up production quickly enough, the industry is estimated to decline at a CAGR of 3.0% through 2024. Demand for aircraft is expected to remain robust, with production gradually increasing, leading to a projected annual industry growth of 2.4% and reaching $361.1 billion by 2029. This strong demand will be driven partly by rising disposable incomes and urbanization in developing markets. Additionally, the e-commerce boom will fuel the need for cargo aircraft as logistics companies expand air freight operations to meet consumer expectations. However, as demand grows from emerging markets, competition from companies like COMAC, Irkut and Embraer will intensify. The increased supply and heightened competition are likely to exert downward pressure on the profit margin.

The HRPlanesv2 dataset contains 2120 VHR Google Earth images. To further improve experiment results, images of airports from many different regions with various uses (civil/military/joint) selected and labeled. A total of 14,335 aircrafts have been labelled. Each image is stored as a ".jpg" file of size 4800 x 2703 pixels and each label is stored as YOLO ".txt" format. Dataset has been split in three parts as 70% train, %20 validation and test. The aircrafts in the images in the train and validation datasets have a percentage of 80 or more in size. Link: https://github.com/dilsadunsal/HRPlanesv2-Data-Set

Description:

👉 Download the dataset here

This dataset is an extensive collection of high-quality images featuring various fighter aircraft, curated specifically for machine learning tasks such as classification, image recognition, and object detection. The images are systematically organized into folders, each labeled with the corresponding aircraft model name, offering a valuable resource for researchers and developers working on AI-driven aviation projects.

Download Dataset

Aircraft Models Included

The dataset includes a wide variety of modern and vintage fighter aircraft from around the world. Examples range from early 20th-century designs to cutting-edge jet fighters, making it suitable for research across various eras of aviation technology.

Examples of Fighter Models:

F-22 Raptor

Su-30MKI

Eurofighter Typhoon

F-16 Falcon

Each model folder contains a comprehensive set of images covering key visual aspects of the aircraft, such as wing design, fuselage structure, and engine placement, allowing for high-precision classification models.

Applications in Machine Learning

This dataset is an essential tool for tasks like:

Object Detection: Training models to detect fighter aircraft in aerial images or videos.

Image Classification: Creating models that can differentiate between various aircraft types based on visual features.

Feature Extraction: Studying unique design features that distinguish different fighter jets, assisting in AI-based image feature extraction.

Dataset Specifications

Format: JPEG/PNG files

Size: The dataset contains thousands of images with varying resolutions to fit different model requirements.

Annotations: Optional annotation files provide bounding boxes and additional metadata for object detection tasks.

Conclusion

The Fighter Aircraft Dataset offers a well-rounded collection of fighter aircraft visuals, enhancing machine learning efforts in both the defense and aviation sectors. Whether you’re building an AI model for image classification or exploring aviation research, this dataset serves as a critical resource for accurate and efficient learning models.

This dataset is sourced from Kaggle.

OWLETS2_UMDAircraft_Data_1 is the Ozone Water-Land Environmental Transition Study (OWLETS-2) data collected onboard the University of Maryland Cessna Aircraft. Data include trace gas measurements, greenhouse gases, aerosols, and aircraft navigational and housekeeping data collected via remote sensing and in-situ instrumentation. This collection features data from the GeoTASO instrument, a pre-cursor to the TEMPO satellite. OWLETS and OWLETS-2 were supported by the NASA Science Innovation Fund (SIF). Data collection is complete.Note: The GeoTASO data included in this collection was collected onboard the HU-25A aircraft.Coastal regions have typically posed a challenge for air quality researchers due to a lack of measurements available over water and water-land boundary transitions. Supported by NASA’s Science Innovation Fund (SIF), the Ozone Water-Land Environmental Transition Study (OWLETS) field campaign examined ozone concentrations and gradients over the Chesapeake Bay from July 5, 2017 – August 3, 2017, with twelve intensive measurement days occurring during this time period. OWLETS utilized a unique combination of instrumentation, including aircraft, TOLNet ozone lidars (NASA Goddard Space Flight Center Tropospheric Ozone Differential Absorption Lidar and NASA Langley Research Center Mobile Ozone Lidar), UAV/drones, ozonesondes, AERONET sun photometers, and mobile and ship-based measurements, to characterize the land-water differences in ozone and other pollutants. Two main research sites were established as part of the campaign: an over-land site at NASA LaRC, and an over-water site at the Chesapeake Bay Bridge Tunnel. These two research sites were established to provide synchronous vertical measurements of meteorology and pollutants over water and over land. In combination with mobile observations between the two sites, pollutant gradients were able to be observed and used to better understand the fundamental processes occurring at the land-water interface. OWLETS-2 was completed from June 6, 2018 – July 6, 2018 in the upper Chesapeake Bay region. Research sites were established at the University of Maryland, Baltimore County (UMBC), Hart Miller Island (HMI), and Howard University Beltsville (HUBV), with HMI representing the over-water location and UMBC and HUBV representing the over-land sites. Similar measurements were carried out to further characterize water-land gradients in the upper Chesapeake Bay. The measurements completed during OWLETS are of importance in enhancing air quality models, and improving future satellite retrievals, particularly, NASA’s Tropospheric Emissions: Monitoring of Pollution, which is scheduled to launch in 2022.