These are articles related to using the Nvidia Jetson TK1 embedded supercomputer. Most of the information on this is going to be associated with how to use the TK1 for robotics projects and interfacing it with stand-alone microcontrollers like the Arduino for sensor and motor integration.
Improved fully convolutional neural network model for locating and identifying plants in a crop field image as either corn, okra, or weeds. This network will be used in a weeding robot. It shows a dramatic improvement in its ability to find plants within the images.
[Read the rest of this article...]
Discuss the initial steps for building a weeding robot that can eliminate weeds in gardens and farms. This blog post describes some of the first test neural networks to help identify weeds within crop fields.
Jetduino prototype boards are ready and the contest to win a free Jetduino robotics interface board for the Jetson TK1 will begin soon!
Discussion of the test results from version 0.3 of the Jetduino robotics interface board for the Jetson TK1.
Strategies for reducing the cost of the Jetduino, and what impacts that will have. Description of the different board options, and estimated prices.
Description of the Jetduion V0.2 test results and future plans.
Describes how to control standard servos on the Jetson TK1 using the Jetduino. A YouTube video is included that show how to do this step-by-step.
Describes how to setup the Jetduino for the Jetson TK1. Then shows how to access digital I/O lines for both the Arduino Due and the Jetson's GPIO. YouTube videos are included that show how to do this step-by-step.
Please subscribe to my Newsletter!
NeuroRobotic Technologies is dedicated to creating the next generation of intelligent, adaptive robotic systems by building autonomous controls systems that mimic the brains of real animals.