Project14 | Winners Announcement:  Vision Things Using Graphics, AI, Computer Vision, and Beyond!

The Winners

The Grand Prize

BBAI Seein' Around Corners, Talkin', IoT Exploitin' Backup Car Cam with Onboard Vision AI  by Sean_Miller:

• BeagleBone AI Survival Guide V3.18: PWM, I2C, Analog/Digital Read/Write, Vision AI, Video Text Overlays, Audio, & Hardware

Sean_Miller  was happy to be provided a BeagleBone AI to experiment with for this Project.  So, he and his son connormiller  scratched their heads hard on what they could do that would be intriguing to a reader while testing nearly all functionality of the BBAI.  The Debian Linux operating system works great with the BBAI and provides many project opportunities on its own.  To take it further they wanted to make something original that would exploit all its capabilities:  GPIO Input/Output, PWM, I2C, Visual AI Classification with TIDL, Open CV for a GUI, Ad Hoc WiFi Streaming, IoT API communication, and Audio.  They also thought of how an educator could use it as a classroom project - dig out those Matchbox cars! So, they dreamed up the Seein' Around Corners, Talkin', IoT Exploitin' BBAI Backup Car Cam.  It's a novelty project intended to provide example code to get one going with the game changing BBAI - not as another Linux box with a web cam - but as a GPU accelerated, embedded micro controller to control your own inspired Vision AI robotics projects. Of course, the BBAI can readily showcase existing Raspberry Pi project repositories.  Their mission was to showcase the BBAI's hardware advantage.  This project leverages TI's Vision AI accelerated hardware to max out the BBAI capability at frame rate speeds.  With the BBAI being brand new hardware to the market, this project led to greater than 90 hours of research, design, coding, and documentation as we went down a rabbit hole to understand the BBAI architecture.  They eliminated much clutter from their blog by documenting the research into a companion blog that is sure to get you going with the BBAI.  Their hope is that you enjoy their virtual backseat buddy and gain some nuggets of your own to make for this great platform.

Community Member Scoring:

Grand Prize: 8 points, First Place:  2 points, Total Points: 10 Points

Seeing Around Corners BeagleBone Backup Car Camera

"He proposed an ambitious project, completed it, and along the way documented important functions in the new BB AI.  Other contestants benefited from his documentation.... I will be referring to his information as I learn the BB AI." - Community Member Judge

"Nice project with a mix of technologies. However it wasn't clear how much of the reversing activity was using video, or the LIDAR sensor. Still, the technologies used were great to see, and the information could be applied to many other future projects. Sean put in a lot of effort, as well as lots of useful BB-AI documentation." - Community Member Judge

Edit media

Dimensions x Small Medium Large Custom

Subject (required) Brief Description Tags (separated by comma) Video visibility in search results Visible Hidden

Parent content

Project14 | Winners Announcement:  Vision Things Using Graphics, AI, Computer Vision, and Beyond! Project14 | Vision Thing: Meet the Projects Using Graphics, AI, Computer Vision, & Beyond! Project14 | Volunteer to Judge the Vision Thing Competition!

Poster

Upload Preview

Ok

Cancel

"Sean Miller has went the extra mile ahead and explored the BB-AI. He has provided so much valuable information that will help a starter have a hassle free start. His project will be definitely useful and has been documented excellently with all the troubles and issues he faced and details of how he overcame them." - Community Member Judge

"My top pick is the "Seein' around corners" project. I really liked reading to its comprehensive project description, and learned quite a lot. Sean and his son set their goal quite high, and I think they even overachieved. Bonus points for using LEGO in their test setup :)" - Community Member Judge

"Seeing around corners, excellent post providing detail, pictures and good explanation." - Community Member Judge

"Great to see a father and son working together. It's always good to inspire the next generation. I'm hoping to do the same when mine are a little older. I also appreciated the fact that a useful guide to getting started with the BB-AI came out of this too." - Community Member Judge

First Place Winners

Vector Display GPU Project  by neuromodulator :

As a follow-up to Learning Verilog with the Digilent Cmod S7  blog post, and as part of the Project14 | Vision Thing: Beaglebone AI Your Vision Thing Project!   neuromodulator decided to use an FPGA to design a GPU to draw on a vector display  Vector displays in contrast to raster displays can have the electron beam deflected in any arbitrary form through the control of their X and Y coordinates. One example of this type of display is the Tektronix 4051.  Another example of displays that use this technology are cathode ray tube (CRT) oscilloscopes. Having the ΔΣ DAC properly working he decided to go for something more challenging. Lissajous curves have already been shown too many times on oscilloscopes, so he decided to solve ordinary differential equations in the FPGA. One interesting ODE system is the Lorenz system, developed by Edward Lorenz while studying atmospheric convection, which can display chaotic behavior  with certain parameter values.

Community Member Scoring:

Grand Prize:  6 points, First Place:  2 points, Total Points: 8 Points

Vector Display GPU Project

"I learned a lot reading this (Q notation), I expected a simple Lissajous vector display (which he noted was overdone).  Very impressive and unexpected project.  I would have to spend many days to understand all that he presented." - Community Member Judge

"A really interesting project that was also very well written up. I liked the fact that it merges of an old style of display with a modern FPGA." - Community Member Judge

Edit media

Dimensions x Small Medium Large Custom

Subject (required) Brief Description Tags (separated by comma) Video visibility in search results Visible Hidden

Parent content

Vector Display GPU Project Project14 | Winners Announcement:  Vision Things Using Graphics, AI, Computer Vision, and Beyond! Project14 | Vision Thing: Meet the Projects Using Graphics, AI, Computer Vision, & Beyond! Project14 | Volunteer to Judge the Vision Thing Competition!

Poster

Upload Preview

Ok

Cancel

"A fantastic project (actually two projects rolled into one). Some source code was provided. Really great to see FPGAs being used for vector displays, and the demonstration video for the second project (which used OpenCV as a kind of preprocessor) was impressive too." - Community Member Judge

"Vector Display, good technical challenge appreciated by those of us experienced with vector graphics design and development." - Community Member Judge

Light Dependant Resistor Camera LDRCam)  by dubbie:

• LDR Camera #2 : Adding an 8x8 Dot Matrix LED Display

• LDR Camera #3 : Trying to make a Grey Scale LED matrix Display

• LDR Camera #4 : The BeagleBone AI has Landed

• LDR Camera #5 : A 64-to-4 Analogue Multiplexer

• LDR Camera #6 : Starting to Implement the Whole Camera System

• LDR Camera #7 : 98% Working

• LDR Camera #8 : Fully Working Camera

• LDR Camera #9 : Starting the BeagleBone AI

• LDR Camera #10 : The Artificial Neural Network (SimonnV5)

  • BeagleBone AI Terminal pins connector

  • BeagleBone AI Terminal Connection

dubbie  couldn't think of any more excuses for not powering up the BeagleBone AI so that he could use it to run the ANN (artificial neural network) for his LDR Camera, He decided to try and get it going. He thought he had a Raspberry Pi 3 power supply that came with a Type C connector but as it turns out, he doesn’t. He also thought he had a power module with a USB connection that was able to provide 5V at 2 amps or so. Turns out he doesn’t, it is 5.2V at 1.8 A and as he wasn't sure if 5.2V would be OK he decided not to use it. All that was left was the USB C cable to his laptop that he had previously purchased. The Artificial Neural Network that dubbie  wants to use with the LDR camera and the BeagleBone AI is one derived from a book by H Pao. As it has been some time since he used this programme he thought it would be beneficial to refresh his memory of how it works. Fortunately he has a copy of the program with will work in a Command window with Windows 10. He no longer has a copy of the Microsoft C compiler so he was not able to make any changes to this version. Hopefully he will be able to use a C compiler on the BeagleBone AI to make some updates. The consequence of this is that the programme, called SimonnV5 (SIMple Online Neural Network Version 5 - it's not online but he needed the O to make Simon) is limited to a maximum of 50 inputs. The LDR camera has 64 inputs so he cannot use it with compatible data examples.

Community Member Scoring:

Grand Prize: First Place:  4 points, Total Points: 4 Points

Light Dependent Resistor Camera (LDRCam)

"His step by step description of his work was excellent, problems encountered, workarounds he decided on.  He started with what he knew (arduino), making his concept work, and then tried to move on to the BB AI.  His explanations and videos show his thought process and results clearly." - Community Member Judge

Edit media

Dimensions x Small Medium Large Custom

Subject (required) Brief Description Tags (separated by comma) Video visibility in search results Visible Hidden

Parent content

Project14 | Winners Announcement:  Vision Things Using Graphics, AI, Computer Vision, and Beyond! Project14 | Vision Thing: Meet the Projects Using Graphics, AI, Computer Vision, & Beyond! Project14 | Volunteer to Judge the Vision Thing Competition!

Poster

Upload Preview

Ok

Cancel

"This turned out to be a very unique project. Initially I wasn't sure what the result would be, but Dubbie managed to turn it into an interesting hardware version of an eye, with a ton of wiring to do that! Impressive that he even went to the extent of investigating a neural network for the eye too!" - Community Member Judge

"LDR Camera, Great post, original idea and execution" - Community Member Judge

Driver State Monitor with OpenCV and BeagleBone AI  by vlasov01:

This project is an attempt to build a in-car safety device, which monitors drivers' attention to the road and creates alerts if the driver becomes sleepy or distracted. The statistics are quite alarming in regard to drowsy driving. An estimated 1 in 25 adult drivers (aged 18 years or older) report having fallen asleep while driving in the previous 30 days in the US. The US National Highway Traffic Safety Administration estimates that drowsy driving was responsible for 72,000 crashes, 44,000 injuries, and 800 deaths in 2013. However, these numbers are underestimated and up to 6,000 fatal crashes each year may be caused by drowsy drivers. Texting while driving is another significant source of accidents a well. Journey Driver State Monitor (DSM) with BeagleBone AI is a project that explores the capabilities of OpenCV and BeagleBone AI (BBAI) to improve drivers' safety. It is based on work done by Adrian Rosebrock, which was published in his blog "Drowsiness detection with OpenCV."

Community Member Scoring:

First Place:  3 points, Total Points: 3 Points

Driver State Monitor with OpenCV and BeagleBone AI

"This was a genuinely useful implementation of the BeagleBone's Vision processing. He saw the project through to completion too. I almost put this as the winner. This and Neuromodulator's display are the two stand-out projects this month." - Community Member Judge

"This project is a life saver I would say. Because most accidents happen in just a blink of an eye of tired drivers." - Community Member Judge

Edit media

Dimensions x Small Medium Large Custom

Subject (required) Brief Description Tags (separated by comma) Video visibility in search results Visible Hidden

Parent content

Project14 | Winners Announcement:  Vision Things Using Graphics, AI, Computer Vision, and Beyond! Project14 | Auto Hacks: Work on Vehicle Projects: From CAN Bus to Camera or Machine Vision Related! Project14 | Vision Thing: Meet the Projects Using Graphics, AI, Computer Vision, & Beyond! Project14 | Volunteer to Judge the Vision Thing Competition! Driver State Monitor with OpenCV and BeagleBone AI

Poster

Upload Preview

Ok

Cancel

"Although the project description is not as complete, Sergey delivered a really cool and useful project here. Using face recognition to detect when someone is about to fall asleep or is not concentrated is a cool idea." - Community Member Judge

"Driver State Monitor, built on someone else's work, but a good use of vision systems." - Community Member Judge

Virtual Loop Sensor  by skywalker1211

This project uses the raspberry pi 2 and raspberry pi  camera v1.3  to record video from a traffic junction/grab it from online and process at home instead of installing raspi on the junction. for video proof, he does a simple vehicle count and detection with mini size vehicle (e.g. remote car) to demonstrate the project output.  For opencv4, you will get plenty of XXX was not declared in this scope errors.  This is due to capture properties no longer begin with CV_ in opencv4, and raspicam is written based on opencv3. workaround is to change all CV_CAP_XXX into cv::CAP_XXX in raspicam source file.

Community Member Scoring:

First Place:  3 points, Total Points: 3 Points

Virtual Loop Sensor

"Pretty impressive work - well documented, and a great concept, to do vehicle sensing by using a line drawn on the video. Great to see OpenCV being used, and the source code will be great for people to follow to understand how to create such video related projects. This was a very complete project." - Community Member Judge

"Virtual Loop Sensor, good idea to set up an advanced vehicle counting package." - Community Member Judge

Runners Up:

Vehicle Speed Camera - Beaglebone AI by gam3t3ch :

Community Member Scoring:

First Place:  1 points, Total Points: 1 Points

Vehicle Speed Camera - Beaglebone AI

"This is a brilliant idea making a speed gun out of BB-AI which could be extended to make use of AI feature to note down frequent or pattern of defaulters by noting down the vehicle registration as well." - Community Member Judge

"There were lots of vehicle projects in this Project14 challenge, but Wesley still managed to keep it interesting, tackling something quite challenging, determining the speed of objects. This could have applications for security purposes too; identifying unusually slow-moving cars, perhaps of the same color or type, multiple times, 'scoping the joint' or neighborhood." - Community Member Judge

Medication Monitor - Part 1 by milosrasic98

Community Member Scoring:

First Place:  1 points, Total Points: 1 Points

Medication Monitor - Part 1

"This is another useful life saver project, not taking medications at the right time could be fatal. Waiting for Part 2." - Community Member Judge

"Awesome idea - very innovative. I think there could be many applications for image recognition built into lighting like this. Milos used OpenCV, and got to the point of detecting the individual compartments of pill boxes, which is very cool. Hopefully the source code can be published too so people can learn the techniques used with OpenCV." - Community Member Judge

Self Driving Car (Automatic Lane Detection) (on a Robot) by vimarsh_

Community Member Scoring:

First Place:  1 points, Total Points: 1 Points

Self Driving Car (Automatic Lane Detection (on a Robot)

"Another well-done projects with a comprehensive project description. Vimarsh did a great example of all the work that needs to done for self-driving cars, and I again learned quite a lot from his project." - Community Member Judge

"Automatic Lane Detection, good project well explained." - Community Member Judge

"A good project. It still needs work to improve it, but considering Vimarsh attempted both robotics and vision, two challenging subjects, he's very brave : ) Really nicely documented too." - Community Member Judge

Dragon Detector by Workshopshed:

Community Member Scoring:

First Place:  1 points, Total Points: 1 Points

Dragon Detector

"This hit my sweet spot for not-so-serious projects. But even dragons and knights prove to be very useful to learn about AI" - Community Member Judge

"Dragon Detector, definitely original though an old recycled project." - Community Member Judge

"Great project showing the power of OpenCV. Full source code too." - Community Member Judge

Smart Security Camera (with Attendance on Google Sheets) by aabhas:

Community Member Scoring:

First Place:  1 points, Total Points: 1 Points

Smart Security Camera (with Attendance on Google Sheets)

"Well-documented project, showing how to use OpenCV for face recognition and automatic logging of the person to a spreadsheet in the cloud. Aabhas went to a lot of effort to document this project, it looks great." - Community Member Judge

"I can see how much enthusiasm has gone into this, and how one of our younger members has really been inspired to put a lot of effort into E14." - Community Member Judge

"Smart Security Camera, well done project with lots of uses." - Community Member Judge

AI_Lock - DIY with BeagleBone AI in four hours by fyaocn:

• First of All, Tame the roaming beast, BBB AI.

Community Member Scoring:

First Place:  1 points, Total Points: 1 Points

aiLock on BeagleBone AI

"Very interesting project, relying on the BB-AI hardware features and TIDL functionality to perform the image recognition. This is new stuff, so it's great to see such projects we can all learn from. There was actually an immense amount of useful information, but needs a video." - Community Member Judge

"While others seemed overwhelmed by the new BB AI, he got right to it and made an electromechanical solution with no complaints.  He explained advanced BB topics such as PRU and TIDL.  I would have to spend a few days learning to understand it all (and I will, later)" - Community Member Judge