Discussion
We have simplified our project goal, namely that, instead of four we will have three satellite stations, and a "V"-shaped
formation instead of a "circular"-shaped formation.
We have successfully implemented the project we specified in Project 4 with some modifications. The Kinect based object tracker fulfils its design specification. It is inexpensive, easy to setup in varied conditions -- indoors and outdoors and it can track a known number of satellite stations reasonably well in real time. The tracker only uses the depth data provided by the IR projector and IR CMOS sensor of the Kinect camera. Therefore, the tracker works in complete darkness.
Unfortunately, we are not able to provide a video demonstration at this moment as all but one nRF24L01 radios broke down. We were given working radios by Neil and in fact, we were able to communicate between two stations successfully. This led us to labour over days to fix non-existing bugs in our radio code (our reasoning was: as we have seen it to work before by our own code, our recent code must be broken). However, to our dismay, we only recently discovered that only one radio works.
We have found out that the driver provided by Neil work just by properly configuring the board pins with which the radio is connected to. Apart from properly configuring (renaming) the SPI pins, IRQ, CE, CSN and properly resetting the radio when the board powers up, almost no modification is necessary at all. We know of many anecdotal stories of (CSC 460/560) groups labouring over radio codes. Indeed, toward the end of the project we laboured over days (breaking our backs) to fix non-existing radio bugs only to discover that those radios are broken.
We have successfully implemented the project we specified in Project 4 with some modifications. The Kinect based object tracker fulfils its design specification. It is inexpensive, easy to setup in varied conditions -- indoors and outdoors and it can track a known number of satellite stations reasonably well in real time. The tracker only uses the depth data provided by the IR projector and IR CMOS sensor of the Kinect camera. Therefore, the tracker works in complete darkness.
Unfortunately, we are not able to provide a video demonstration at this moment as all but one nRF24L01 radios broke down. We were given working radios by Neil and in fact, we were able to communicate between two stations successfully. This led us to labour over days to fix non-existing bugs in our radio code (our reasoning was: as we have seen it to work before by our own code, our recent code must be broken). However, to our dismay, we only recently discovered that only one radio works.
We have found out that the driver provided by Neil work just by properly configuring the board pins with which the radio is connected to. Apart from properly configuring (renaming) the SPI pins, IRQ, CE, CSN and properly resetting the radio when the board powers up, almost no modification is necessary at all. We know of many anecdotal stories of (CSC 460/560) groups labouring over radio codes. Indeed, toward the end of the project we laboured over days (breaking our backs) to fix non-existing radio bugs only to discover that those radios are broken.