Wednesday, March 30, 2022

T265 Intel RealSense Tracking Camera

 Intel® RealSense™ Tracking Camera T265 is a tracking capable peripheral device based on visual and inertial sensor fusion. The assembly contains fisheye cameras, IMU module and a processing ASIC (Intel® Movidius™ Myriad™ 2 MA215x) with USB 3.0 interface to host processor SoC.

The Intel® RealSense™ Tracking Camera T265 includes two fisheye lens sensors, an IMU and an Intel® Movidius™ Myriad™ 2 VPU. All of the V‑SLAM algorithms run directly on the VPU, allowing for very low latency and extremely efficient power consumption. The T265 has been extensively tested and validated for performance, providing under 1% closed loop drift under intended use conditions. It also offers sub 6ms latency between movement and reflection of movement in the pose. This is fast enough for even highly‑sensitive applications such as Augmented and Virtual Reality.
- Tracking feature using Fisheye Camera and Inertial
Measurement Unit (IMU)
- Intel® Movidius™ Myriad™ 2 ASIC solution
- Middleware processed on Myriad 2 ASIC; Enabling
higher CPU performance
- 6DoF data streaming to host
- Low latency
Spec :
V‑SLAM, part of Intel® RealSense™ Technology
High precision Visual Inertial Odometry Simultaneous Localization and Mapping algorithms.
Intel® Movidius™ Myriad™ 2.0 VPU
Visual Processing Unit optimized to run V‑SLAM at low power.
Two Fisheye lenses with combined 163±5° FOV
The camera includes two OV9282 imagers with fisheye lenses for a combined, close to hemispherical 163±5° field of view for robust tracking even with fast motion.
Inertial Measurement Unit (IMU) :
The Inertial Measurement Unit, allows for accurate measurement of rotation and acceleration of the device, to feed into the V-SLAM algorithms.
USB 3.1 Gen 1 Micro B
USB 2.0 and USB 3.1 supported for either pure pose data or a combination of pose and images.
Dimensions : 108 mm x 24.5 mm x 12.5 mm
Mounting : 2 x M3 0.5mm pitch mounting sockets

Tuesday, March 29, 2022

DuckieTown Invention Kit Self Driving Car powered NVIDIA Jetson Nano

 The DuckieTown founder's edition is a do-it-yourself hardware for assembling a model self-driving car. It is available in 4 variants, defined by the presence of the computational unit and battery. It is designed to operate in a DuckieTown urban environment. The DB has two wheels which can be controlled independently through DC motors. A third passive omnidirectional wheel provides equilibrium. This configuration is known as differential-drive. It is equipped with 4 RGB addressable LED's, plug and play sensor suite, NVIDIA Jetson Nano 2GB developer board and custom made power bank. To facilitate user interface it is also equipped with bi-chromatic I2C screen and a non-latching shut-off button.


  • CPU quad-core ARM A57 at 1.43 GHz
  • Memory 2 GB 64-bit LPDDR4 25.6 GB/s
  • Storage microSD
  • Connectivity Gigabit Ethernet
  • Camera 1x MIPI CSI-2 connector
  • Display HDMI
  • Battery capacity 10000mAh at 3.7V
  • USB ports include 1x USB 3.0 Type A, 2x USB 2.0 Type A, 1x USB 2.0 Micro-B