USB 3.0 - Extending the use of Kinect™ camera
In many Machine Vision applications that use cameras and image processing software, a secured connection is a serious concern since data transfer can slow down over long distances. In early October 2015, Dr. Daniel Lau from the University of Kentucky contacted Newnex for a solution to extend the connection for his research application.
The research aims to show that an automated depth camera machine vision based system has the utilized in respiration rate and stance monitoring of individual dairy cow. In terms of camera, the Microsoft™ Kinect™ is a unique sensor that capable of recording depth data, RGB color information, infrared imaging data, and audio data. In order to safely monitoring the cow and collecting the data, the placement of the host computer should not be place in an operating dairy farm environment. Nonetheless, the use of Kinect™ camera require an USB 3.0 cable connection, and it has a major drawback in which it has a practical cable length limit of 3 meters. The major challenge is to extend the USB 3.0 transmission distance up to 60 meter between the Kinect™ camera and the host computer as figure 1 shown.
Figure 1: For a safe placement of host computer, it will require 60m distance away from the operating dairy farm.
Currently, there are a few other solutions available to extend the USB 3.0 transmission distance such as a passive cable with large wire gauge, an active repeater cable, or an AOC (active optical cable) to extend USB 3.0 signal. Nevertheless, each approach has its own disadvantages. A large wire gauge passive cable with a thicker cable OD (outer diameter) is bulky and less flexible, which makes it difficult for cable management and installation. On the other hand, drawbacks of active repeater cables include a length limit of 20 meters or less in addition to its inability to be EMI shielded. However, these existing solutions fail to resolve the major challenge, where 60 meter distance is required between host computer location and the research data collection location called for the Kinect™ camera to be connected remotely to the computer.
To Break the limited cable length barrier of the USB 3.0, Dr. Daniel Lau apply a set of FireNEX-5000H™ - the world’s first optical repeater that is capable of extending USB 3.0 signal up to 300 meters with 5 Gbps Super Speed bandwidth. While other USB 3.0 optical repeaters currently on the market are unable to minimize the data transmission latency to satisfy the isochronous data transfer’s latency requirement, the FireNEX-5000H™ can upgrades the signal amplification mechanism by optimizing the circuitry design in the physical layer level. With the innovative design that can significantly reduces the USB data transmission latency compared to other USB 3.0 optical repeaters available, FireNEX-5000H™ makes it possible to extend the connection of Kinect™ camera not only with USB 3.0 bulk traffic but also with USB 3.0 isochronous data transfer up to 60 meters. In terms of setup (figure 2), the FireNEX-5000H™ use optical fiber to transmit data, making the installation and cable management significantly easier due to its flexibility and lightweight.
Figure 2: FireNEX-5000H™ System Setup Diagram
Since the overall performance of the monitoring system is highly dependent on the quality of the images that are captured by the Kinect™ camera, Dr. Daniel Lau was pleased with the result as it was the only solution that allowed him to extend USB 3.0 signal to his desired distance up to 60 meters with low latency. With the reliability of the FireNEX-5000H™, he has successfully monitored both the respiration rate and stance of individual dairy cow in real time streaming and recording (Figure 3, 4&5). With the success of the project, this monitoring system will be upgraded so that it can maximizing individual animal potential, detecting disease earlier and minimizing the use of medication in the dairy industry. Furthermore, this research will help farm management to better facilitate resources, monetary and otherwise in an appropriate and efficient manner.
Figure 3: Monitoring system viewed at an angle from the cow with the Kinect™ Camera recording above the cow.
Figure 4: Kinect studio in Color and Grey Point Cloud image mode showing RGB data steam and depth frame respectively.
Figure 5: Kinect studio in Infrared and surface with Normal image mode.
Shelley, Anthony N., "INCORPORATING MACHINE VISION IN PRECISION DAIRY FARMING TECHNOLOGIES" (2016). Theses and Dissertations--Electrical andComputer Engineering. Paper 86. http://uknowledge.uky.edu/ece_etds/86