JeVois is a small, open-source, smart machine vision video camera that was funded on Kickstarter in early 2017. I backed it since cameras that embed device vision aspects are steadily growing much more capable, as well as JeVois boasts an outstanding variety of features. It runs embedded Linux as well as can process video at high frame rates utilizing OpenCV algorithms. It can run standalone, or as a USB video camera streaming raw or pre-processed video to a hold computer for even more action. In either situation it can interact to (and be managed by) other gadgets through serial port.
But none of that is what truly struck me about the video camera when I got my unit. What truly stood out was the demo mode. The team behind JeVois nailed an efficient demo mode for a complex device. That didn’t occur by accident, as well as the results are worth sharing.
The significance of a great Demo
When it concerns complex systems, a great demo mode is essentially an elevator pitch for the unit’s capabilities. To a user, it answers “what can this do, as well as what possibilities does it open for me?”
The JeVois camera’s demo mode succeeded in this by doing a number of things:
Make the demo self-contained as well as simple to start. need a minimum of parts or configuration from the individual to get started. After putting the system picture onto the included SD card, I only needed to plug it in to my laptop as well as begin a camera viewer.
Make it interactive. respond to individual input immediately, as well as show the processes at work as much as possible.
Keep it simple. The demo isn’t the device’s one as well as only chance to discuss everything! Leave the individual totally free to focus on taking in what is being shown; prevent bogging the individual down with figuring out an interface or troubleshooting issues.
Demo mode on hardware is often an afterthought if it exists at all, however it is worthy of interest as well as polish if for no other reason than it is the one element of a product that it is practically specific every individual will engage with.
Setup as well as Demo of JeVois
I had to copy a system picture to the micro SD card to guarantee I had the most recent version of the JeVois software, however after that all I needed to do was plug it into a USB port (where my computer acknowledged it as a USB video camera device) as well as open a webcam viewer. when that was done, the demo immediately started. The hardware streams a hybrid blend of a looping “tour” video mixed with online video camera pictures as well as a number of vision processes running in parallel. To the hold OS, it’s just a video stream from a USB video camera like any type of other.
Not every aspect is flawless; the handwritten number recognition in specific is hit or miss, although the demo acknowledges this. The demo video below was recorded in a white space with sunlight, however things still look on the dim side. Finally, the video camera as well as lens clearly aren’t meant to be altered or adjusted. I’d have favored a module with an M12 (S-mount) lens, however the video camera on JeVois at least makes for a extremely little package. Regardless, the demo succeeds in easily as well as concisely showing off what’s possible.
The default video camera system picture includes a range of bundled device vision modules as well as demos, each anchored to a particular video camera setting. For example, one of these modules recognizes ArUco markers. To activate this mode, set the video camera to “YUV 640×500” as well as the module activates immediately. A recording of this mode is below.
Details on detected markers are sent over the serial port. In this method an Arduino or microcontroller can interface to as well as even manage the camera. There are numerous other bundled modes in the default image, as well as of program individuals are totally free to establish their own.
Other Embedded device Vision Cameras
[image source: CMUcam.org]
[image source: OpenMV]
The CMUcam5 “Pixy” (also funded on Kickstarter) shipped in 2014 had a smart demo mode. Out of the box, it had the capability to acknowledge colors that were held in front of it, as well as even had headers for pastime servos to be utilized in a pan/tilt configuration. This enabled the video camera to be trained to track a color, then relocation to comply with that color, all without including a computer. It was likewise possible to view raw or processed video, however that needed linking to a laptop or desktop by USB as well as running a program distinct to the Pixy in purchase to gain access to video as well as configuration. The capability to “see what the video camera sees” was a fantastic function that made configuration as well as troubleshooting much easier, even if it was limited.
The OpenMV project is an open source python-enabled device vision module that was a semifinalist for the Hackaday Prize in 2014, as well as has grown because then. OpenMV have just started shipping their newest M7 cameras, which can even emulate a Pixy in UART mode for compatibility. We’d like to see a great demo to show off its capabilities.
Are there any type of other new offerings in the realm of embedded device vision cameras, or especially great demos? let us understand in the comments. as well as speaking of the Hackaday Prize, don’t fail to remember that 2017’s iteration is currently underway.