|
|
|
|
|
|
|
KudanSLAM
Artificial Perception algorithms
for all the embedded devices
Kudan, Inc. has succeeded to develop real time 3D mapping and position tracking via camera, called “KudanSLAM*1”, and started to provide its technology to the market such as Autonomous car, ADAS*2, Drone, Industrial and Personal Robots in addition to the existing AR/VR industries.
SLAM, is the software technology, which is capable of 3D mapping and position tracking. It provides computers the ability of “computer vision” to acquire, process, analyse and understand digital images as well as the ability to map its 3D environment, objects, and understand its location within it. This “Computer Vision” technology can be used for any industries such as Autonomous car and Robotics.
Kudan has been developing tracking space and object technology through AR. As a result, Kudan succeeded to develop practicable and next generation algorithm, which would replace the existing SLAM such as ORB and PTAM*3 SLAM base, and apply those technology to be ready for the market.
Kudan, as a SLAM’s leading company, aims to spread use of KudanSLAM which is to be embedded on all image-related devices with camera, in any fields such as Autonomous car, ADAS, Drone and Robotics in addition to the existing AR/VR area.
KudanSLAM has been designed from the ground up to be as flexible as possible, allowing it to be applied in a large variety of circumstances.
The core codebase can target most processor architectures, and there is no reliance on the presence of operating system functionality. Multiple processor classes can be utilised, ranging from low-powered general purpose to highly custom DSPs. A large variety of hardware sensors are supported, ranging from monocular and stereo cameras, up to visual-inertial depth cameras.
Our SLAM is designed to be as general purpose as possible. It can be used equally well in a variety of situations, ranging from mobile positional tracking through to autonomous driving.
Every aspect of the system is highly configurable and exposed via a simple to use API, allowing easy tuning to the target hardware and use-case.
KudanSLAM is designed to be highly performant; computationally efficient as well as providing robust, accurate results.
Everything from algorithm design through to implementation targets speed. This allows for low-latency tracking at high framerates, all while minimising CPU utilisation to save power. Lower powered classes of processor can easily be used.
A unique approach to tracking and mapping provides the most accurate positional and pointcloud data and keeps drift under control. High precision tracking maintains a low amount of jitter.
The real world isn’t always as ideal as in datasets, so we make sure to test in difficult conditions such as bad lighting and fast camera motion.
There are lots of building blocks to a SLAM system, and sometimes there can be a benefit to assembling them differently. We’ve accumulated a lot of these building blocks over the years and provide them as individual modules, allowing their many benefits to brought outside of our SLAM system.
Different approaches to tracking and mapping are available, as well as modules such as loop detection and closure, relocalisation and bundle adjustment.
Efficient implementations of various point matching mechanisms, stereo matchers and pose estimation. All are highly configurable and optimised, utilising better algorithms unavailable in the public domain.
Highly optimised versions of common vision processing building blocks such as various blurs, interpolations and image warps. These are typically SIMD optimised and provide far superior performance compared to OpenCV. We also provide our own linear algebra library.
While our focus is on SLAM itself, it’s often useful to have different modules to help with integration. We provide a GUI library designed for cross-platform computer vision debugging and as well as well as modules to work with the generated SLAM maps.
© Copyright 2000-2023 COGITO SOFTWARE CO.,LTD. All rights reserved
