Skip to main content Skip to first level navigation
Skip to main content First level navigation Menu
Breadcrumbs

Share

Main content

Autonomous flight and obstacle avoidance of drones based on 3D vision system

Autonomous flight and obstacle avoidance in a non-GPS environment using Ricoh eyes

Background

Drones are being used in many industries: structure inspection, surveillance, measurement, and logistics.

Though drones are used across many industries, the way they are used is still very restricted. Basically, their flight is controlled manually under visual surveillance. Their autonomous flight is limited to areas where GPS (global positioning system) is available. Where GPS signals are unstable or unavailable – for instance, under a bridge, in a tunnel, or indoors – or where many obstacles are in the way, technological hurdles remain.

Solutions

To enable drones to be used indoors and where GPS signals are unstable, an alternative to GPS is required. To fly autonomously without GPS, the drones will have to identify their positions on their own. They will also need to detect and avoid obstacles. Ricoh has applied its 3D vision system based on an ultrawide-angle stereo camera to drones, enabling them to fly autonomously indoors where no GPS signals are available, and to avoid obstacles.

Technical highlights

The RICOH SV-M-S1 stereo camera for 3D measurement, which features high speed, high precision, and versatility, has been around for some time. It is used across a range of systems-robot vision, surveillance, object recognition, and more.

The technologies used in the SV-M-S1 have been applied to the super-wide stereo camera for drones. The camera enables autonomous flight and avoidance of obstacles where GPS signals are not available.

With the two features listed below, the drone can generate a wider 3D map than a conventional stereo camera. The drone can quickly acquire the space information required for its flight, including its own position and a 3D map of the areas around the flight route.

● Wide view from a super-wide-angle stereo camera
● Real-time processing of stereo depth measurement, self-localization, and 3D map generation

stereoscop,distance measurement,self-localization

The University of Tokyo, Blue Innovation, and Ricoh have jointly developed an autonomous system for flight and obstacle avoidance by combining outputs from the super-wide-angle stereo camera with those from an inertial navigation system (IMU sensor*1).

*1: IMU sensor: A sensor that measures the motion of a moving object in a three-dimensional space using an accelerometer and gyro.

Ricoh's vision

Ricoh will continue its research and development to create a comfortable society where people and drones coexist in security and produce value together.

*This research is a joint effort between Ricoh, the University of Tokyo, and Blue Innovation.