Login/Register
Hint: separate multiple terms with commas
Date
E.g., 06/24/2019
Date
E.g., 06/24/2019
Feedback
Global
Global
China (简体中文)
Japan (日本語)
Korea (한국어)
Taiwan (繁體中文)
TI Training & Videos >
Applications & designs
Live training
Products
Tools & software
Toggle navigation
Hint: separate multiple terms with commas
Date
E.g., 06/24/2019
Date
E.g., 06/24/2019
Global
China (简体中文)
Japan (日本語)
Korea (한국어)
Taiwan (繁體中文)
Email
3.1 Comparison of 3D ToF and mmWave for Robot Vision (Part 3)
Hello, everyone. This is Richard Wang from Industrial System team of Texas Instruments. Welcome to part three of the training series on comparison of 3D time of flight and mmWave for robot vision applications. In this training, we will see the application scenarios in robot vision for 3D time of flight and mmWave sensor. First, the robot navigation. SLAM, simultaneous localization and mapping, is the key for robot navigation. Putting a robot in unknown position [INAUDIBLE] unknown environment, for example a new room, that it should have the capability of building a map of a room by scanning and localize its real time position to the map. This video is a demo for mmWave to build a map of a room. MmWave sensors measure both the range and angle by illuminating the environment with RF chirp. MmWave sensor measure both the range and angle by measuring the environment with RF chirp. By using range and angle, a coarse map of the environment can be constructed. And then with the map and the motion inputs localization can also be achieved. This video shows a robot able to build up a map based on 3D time of flight. 3D time of flight sensor provides 3D depth data of a room. Supporting visual detection of landmarks needed for [INAUDIBLE] teaching. With that, robot compute 3D maps of multiple rooms. [INAUDIBLE] and visual [INAUDIBLE] maps can be optimally combined by [INAUDIBLE] for localization. Oftentimes, a robot needs to detect a small object in front of it. For example a [INAUDIBLE] robot should have the capability to detect wires, chair legs, shoes on the floor, and avoid them. This video is a demonstration for [INAUDIBLE] robot equipped with a TI mmWave sensor to detect obstacles in front of it and plan to pass to avoid. MmWave will be able to detect slim objects, especially if they are wooden, metal, or other dense material. [INAUDIBLE] we have placed a variety of glass, water, and metal obstacles on the way. The millimeter wave stands provides the position of these obstacles and the robot can plan to pass to the other end of the table. And this video shows how 3D time of flight performs at obstacle detection and avoidance will work in robot application. Based on [INAUDIBLE] 3D time of flight, the robot will slow down when approaching [INAUDIBLE] obstacles. Sometimes the obstacles are small and the thing considers show the [INAUDIBLE] wires and chair legs that can jam up [INAUDIBLE] robot. 3D time of flight can help robot escape ability. Rearranging some limited space, the robot can also cover and escape. For [INAUDIBLE] gesture recognition, both technologies can realize some specific gestures. This video shows gesture recognition by mmWave radar. We can use data output from IWR443 to recognize a gesture. Unique [INAUDIBLE] for mmWave sensor can be used to determine the rotation direction of the finger. For example, zooming up and zooming down. This demonstration is even possible through materials just like glass or dry wood. 3D time of flight can distinguish a hand from a background. And by monitoring the hand shape, gesture recognition can be realized. And this is a automated gesture demonstration. 3D ToF sensor can recognize numbers by hand gesture. Most of the technology can be used for people counting or occupancy detection. This video shows people counting by a 3D time of flight. Please pay attention to the count here. Where there are only one people in the room, the count number is one and the number changes from time to time based on the extra number of people in this area. This video shows the promise of mmWave radar on people counting. The radar detects and tracks the moving target in a scene. For harsh environment applications, for example operating under fog and the rain, mmWave would be a better choice. This video simulates the rain, and we can see that mmWave can detect a target easily as shown here at peak range [INAUDIBLE]. Robot needs to plan paths properly instead of randomly. Optimize a cleaning pass to minimize overlap and maximize the surface area covered. Localize and recognizes the charger and plan paths back for recharging. MmWave sensor can create a map of the environment that can be used by the algorithm to figure out best paths. The sensor will update the environment map, but it move in real time. 3D ToF high resolution sensors help robot build accurate 3D maps of the rooms and localize its position. Robot can plan paths easily based on the map and the localization. Even if the charger is moving away the 3D ToF sensor can help robot to recognize it by a 3D model of the charger. High pixel count and depths enable automatic alignment with contacts for docking and self-recharging. When approaching into a limited space, robot should have the capability of estimating if the space is enough to go through or else get stuck. MmWave sensor can map a 3D space but angle resolution may not be accurate enough to detect very tight areas which is slightly larger than the robot. For 3D ToF, high resolution in short range is sufficient to measure the space precisely without being stuck. For some special cases, for example in front of obstacles like a mirror or glasses, mmWave would be a better choice. Here is a range and accuracy comparison of these two technologies. For example, mmWave can detect the range of 175 meters with less than ten centimeters accuracy. Ten meters was less than one millimeter accuracy, and two point five meters less than 15 micrometers. For 3D time of flight, spatial resolution dependent on FOV. For example, OPT8231 with a 74 FOV lens has a special resolution for six millimeters for object in 1 meter away. The ToF's resolution is 3.5% to 1% of the range.
Watch the previous video in the series:
Comparison of 3D ToF and mmWave for Robot Vision (Part 2)
Details
Date:
April 27, 2018
This is the third video discussing 3D ToF and mmWave for Robot Vision.
Additional information
To learn more, visit our landing page here.
Related courses and events
3D time-of-flight sensors
Date:
August 31, 2016
3D TOF robotics: obstacle detection, collision avoidance and navigation
Date:
September 13, 2016
Recommended design process and resources
Date:
October 25, 2016
Application deep dive: Robot navigation
Date:
October 27, 2016
Introduction to Battery Powered Motor Drive Applications
Date:
November 3, 2016
arrow-top
close
delete
download
menu
search
sortingArrows
zoom-in
zoom-out
arrow-down
arrow-up
arrowCircle-left
arrowCircle-right
blockDiagram
calculator
calendar
chatBubble-double
chatBubble-person
chatBubble-single
checkmark-circle
chevron-down
chevron-left
chevron-right
chevron-up
chip
clipboard
close-circle
crossReference
dash
document-generic
document-pdfAcrobat
document-web
evaluationModule
globe
historyClock
info-circle
list
lock
mail
myTI
onlineDataSheet
person
phone
question-circle
referenceDesign
shoppingCart
star
tools
videos
warning
wiki