About This ProjectWe are the first to develop a drone-based robotic lighting system for photography. A quadrotor drone carrying a light source (flash strobe) can automatically position itself with respect to the subject and the photographer to provide perfect lighting. Our overarching goal is to make our drone based lighting technology easily accessible to the photography community. We want to take our in-lab prototype system and adapt it for outdoor photography.
Ask the ScientistsJoin The Discussion
What is the context of this research?
Lighting is crucial to the art of photography. Lights are cumbersome to setup and often impossible to place because of several constraints. Once the photographer chooses a certain theme and lighting style, the placement of the light becomes a fairly mechanical process. This task is perfect for a robotic system. We use drone as a light carrying platform because it can fly, hover and react quickly.
Our recent paper Computational Rim Illumination Using Aerial Robots, Manohar Srikanth (MIT), Kavita Bala (Cornell), Fredo Durand (MIT) International Symposium on Computational Aesthetics, Aug 2014 demonstrated a robotically assisted lighting system to help photographer achieve rim-lighting. We chose rim-lighting as an example, but this approach can be extended to a wide range of lighting styles.
What is the significance of this project?
This project will helps us translate our research prototype into practical tools for photographers, both professionals and non-professionals alike. Accomplishing this requires thorough engineering analysis and study of lighting ergonomics in realistic shooting scenarios. Because drones are limited in their capabilities (pay load capacity, flight time, response time, so on), optimal utilization is the key. Our findings will help photographers design their own basic drone lighting system using off-the-shelf components.
Algorithms are key in achieving automatic lighting. Outdoor experimental trials will help us design sophisticated and robust algorithms for lighting optimization and drone control. Advances in lighting algorithms can have far reaching impact on photographic lighting in general.
What are the goals of the project?
- Perform thorough engineering analysis of the system and develop a set of design guidelines
- Design a new drone lighting platform with off-the-shelf components: with do-it-yourself approach in mind
- Perform several trial runs using manual flight control mode and evaluate platform performance
- Generate rich data (flight log + photographs + video feed) that can be used for off-line analysis and development
- Understand and document lighting ergonomics and document wish-list for in-camera user interface
- Document feedback from the photographer and the model
Like any other technology projects, our project requires use of expensive equipment. We have used our personal funds to procure most of the camera and photography gear (worth well over $25k). We will use Experiment.com funding to procure additional equipment and also to cover crew expenses.
Using the new equipment (drones, depth sensors, etc), we will be able to perform several hours of flight test on multiple photo shoots. This will not only help us gain valuable insight into the lighting ergonomics, but also help us test various algorithms. The data recorded during these trials will help us perform offline analysis, which will in turn help develop advanced lighting algorithms.
Meet the Team
Manohar Srikanth graduated from MIT with a PhD. He is an inter-disciplinary researcher whose research is at the intersection of computational photography and robotics. www.msrik.com, www.photonish.com
Kavita Bala is a professor of computer science at Cornell Univ.
Fredo Durand is a professor of computer science at MIT.
Additional InformationOur first attempt at outdoor trial using our personal funds: http://youtu.be/2R6AEDONbk8?list=UUW0RlBgt-uRepAOz... More photographic results can be found at www.photonish.com
- $586Total Donations
- $36.63Average Donation