I have developed a system to call a robot from remote locations by taking a picture of the current position using a mobile phone camera. Processing contents is as follows; The captured image is transferred to the robot automatically, where place recognition using the image is performed. The recognition process is performed by describing the features as binary code with BRIEF descriptors for the FAST key points, and matching the image to template images based on the Hamming distance. We have also attempted SIFT, BRISK, ORB, and FREAK for key points and descriptors. The robot performs path planning to the location (position) that is stored with the matched template image on a map generated by SLAM in advance, and travels by autonomous navigation.
Carry the robot is troublesome! I want the robot to come with an automatic! This was my wish and an opportunity for the robot is made. This robot to follow the person. He is change the face by the actions of person. This pattern is as follows: 1. If I walked, to follow. 2. If I stopped, He inflames. 3. If the He was away from me, He increases the speed. 4. After that, when he caught up with me, he to shortness of breath. 5. If I sat down, he comes up to me. 6. If he pass each other, to he says hello. I was devised to avoid false detection and malfunction. For example, he don't see the wall, and he follows me straight.
It is difficult to make the software for robot systems, because robot systems usually require various function. Therefore, the use of someone's software like open source is important. This demonstration introduces my software for autonomous navigation. It supply with mobile robot localization, and apply to be easy to make mobile robot system. Using this software, I made a post officer robot.
In this demo program, the robot bring a chair to a master who don't want to go no farther. The robot is tracking the master's position and waiting for a sign. When the master clap hands, the robot brings a chair automatically.
The ability to move unimpeded is of vital importance to our daily lives, and over the years a number of robotics projects have aimed to help ensure the elderly and disabled their mobility. JoLT (a backronym for "Jack of aLl inTerfaces") demonstrates some of the technologies usually integrated to robotic wheelchairs. The system enables a Yamabiko robot to be steered through simple voice commands such as "forward", "turn left", "stop", etc. It also implements automatic obstacle avoidance, the ability to follow a selected beacon, and move through a path delimited by a sequence of visual landmarks.