Pack 19 and other packs from Hopewell District had a beautiful day for fishing and other fun in the sun. This is our Fishing Challenge: North Hopewell Cub Scout Kick-Off to Summer event. The fish were very “smart” with water clear enough that allowed the boys to watch the fish. In the picture below this is Elijah wondering why the fish aren’t biting even though he has dropped the worm right in front of them. Ironically in one case a fish did bite the worm and took the half with out the hook. Another fish took the worm and the hook but the hook came off the line and the fish got away. It was catch and release anyway. The boys had a great time.
Being from Cincinnati I am a huge Cincinnati Bengals fan. No matter how the team performs I am cheering for them usually watching all the games. Tonight I got to meet Coach Marvin Lewis because Elijah has been working hard in school and got on the honor roll. Coach Lewis runs the Marvin Lewis Learning is Cool program so we got free tickets to the Zoo and coupons with Gold Star. While we were at the Zoo we stood in line and got to meet Coach Lewis and get this great picture:
Thanks Coach Lewis!!!! It is incredible that you sponsor such a cool event and encourage kids to do well in school.
Welcome to rovitotv’s world. rovitotv = Todd V. Rovito. It is my Unix username and has been for a long time. I am moving my web site from Google Apps to my own web server so give me some time to move content over. My web site is powered by the Raspberry PI an amazing computer for $40.
For the last two years my Son has been involved in Cub Scouts and Pack 19 in Middletown Ohio puts on a heck of a Pinewood Derby. Pack 19 allows leaders and parents to build their own cars with the hope that it will distract the parents and let the Cub Scouts build their own car. This policy works!!!!! I am one of those Dad’s that obsesses over the Pinewood Derby race. For the 2012 race which was my Son’s first year I generated a good idea to create an air powered car that was to use a rubber band but I didn’t have time to complete it. The 2013 Derby was the debut year of the Speedy Air Powered Pinewood Derby Car!!!! Design changes were made after speaking with a few engineers at AFRL over lunch. The use of radio controlled (RC) parts made this project possible and fairly easy to construct.
In the picture above you can see the car being constructed. We used the following parts:
- Two blocks of pinewood from the standard kit. One block was the car body and the other block was used to build an motor mount
- R/C Speed Controller
- R/C Receiver
- R/C Motor
- Not shown in the picture above battery
- Spektrum DX6i transmitter
- 3 inch 2/3 propeller (buy a bunch of these because a few of these will be broken during testing and construction)
After construction and a few test runs and tweaking the car is shown in the picture below in its transportation box:
Yes I used rubber bands to hold the car together. This is a bad idea especially for our track which doesn’t have the best stop mechanism. A few times during the race I had to realign the parts and apply the rubber bands again. Below is a picture of all the cars before the start of the race, the Speedy Air Powered Pinewood Derby car is in the front right of the picture. I was really worried about the bunny car because it looked fast.
Not shown on the table above another powered car made a appearance at the race (below) but this Dad had solder problems so we didn’t get to race until the end. Please take note I am not the only parent obsessed with the Pinewood Derby.
Below are some bad movies. It was difficult to control the throttle and take movies on my iPhone at the same time. Needless to say this car was very fast. I rarely had to apply much throttle to win. The car weighed in at exactly 5 oz and was the exact size and dimensions as required. It was fun to build and even more fun to race. Thanks to all the people that pitched in I couldn’t of done it without your help and encouragement!
Abstract—This paper will look at using open source tools (Blender, LuxRender, and Python) to generate a large data set to be used to train an object recognition system. The model produces camera position, camera attitude, and synthetic camera data that can be used for exploitation purposes. We focus on electro-optical (EO) visible sensors to simplify the rendering but this work could be extended to use other rendering tools that support different modalities. The key idea of this paper is to provide an architecture to produce synthetic training data which is modular in design and constructed on open-source off-the-shelf software yielding a physics accurate virtual model of the object we want to recognize. For this paper the objects we are focused on are civilian vehicles. This architecture shows how leveraging existing open-source software allows for practical training of Electro-Optical object recognition algorithms.
Links to Paper, Presentation, and Videos:
Abstract – Key idea of this work is to provide a open source frame work for image processing that can be run on different parallel processing platforms. Parallel processing is necessary because sensors are producing more data at faster rates. This presentation looks at wide area motion imagery (WAMI) data from LAIR public released data set. Traditional S&E tools don’t scale in the cloud in terms of performance or cost. Open source software along with cloud computing becomes an enabler of affordable parallel processing because license costs per node are minimized plus computation has become a commodity. This presentation uses Enthought Python Distribution with parallel processing support via built-in ZeroMQ. This process can be automated to produce large amounts of data with minimal human effort.
Links to presentation and sample code:
Abstract – Object recognition is an important problem that has many applications that are of interest to the United States Air Force (USAF). Recently the USAF released its update to Technology Horizons, a report that is designed to guide the science and technology direction of the Air Force. Technology Horizons specifically calls out for the need to use autonomous systems in essentially all aspects of Air Force operations . Object recognition is a key enabler to autonomous exploitation of intelligence, surveillance, and reconnaissance (ISR) data which might make the automatic searching of millions of hours of video practical. In particular this paper focuses on vehicle recognition with Lowe’s Scale-invariant feature transform (SIFT) using a model database that was generated with semi-synthetic data. To create the model database we used a desktop laser scanner to create a high resolution 3D facet model. Then the 3D facet model was imported into LuxRender, a physics accurate ray tracing tool, and several views were rendered to create a model database. SIFT was selected because the algorithm is invariant to scale, noise, and illumination making it possible to create a model database of only a hundred original viewing locations which keeps the size of the model database reasonable.
After a few years of other work this is my re-entry back into pattern recognition.
Links to paper and presentation:
The main focus of this project was to provide an architecture for layered sensing simulation. It is modular in design and constructed on open-source-off-the-shelf-software (Blender, LuxRender, Python, OpenCV, and SUMO).
The project was complied from previous work, fixing many errors such as, multiple vehicle models and color projections. I teamed with and mentored two undergraduate students throughout the process. Their effort enabled them to increase their knowledge and experience with Linux, Python, Ray Tracing, or Geo-Projections.
Links to presentation and video:
When I was younger I was an avid swimmer. My Coach and Friend Mike “Noodles” Nocheck died 6/22/2012 below is an article I found in my archives from the early 1990′s about Noodles and his “retirement” after 16 years at Gamble Nippert YMCA. Did he actually retire? No he went on to coach my high school swim team at Oak Hills for another 20 years. That is over 36 years of coaching!!! Noodles had an intensity about him that brought out the very best in his swimmers. Just when you thought you couldn’t swim another yard he would show you how. Noodles had a connection with his athletes that is difficult to describe, he never cuddled us yet was very good at bringing out the best and making us a little better with each swim. Before each meet we would huddle together and say a prayer, Noodles taught us how to be reverent. I think of all the impurities in sports today and rarely do you hear about a coach as dedicated as Noodles. Surely the world lost a legend on 6/22/2012. All of us that had the privilege to have Noodles as a coach can remember him best by tying to emulate his nature when dealing with kids.
Noodles I will always remember to keep my left elbow up and not drag it in the water……you should take it easy now and retire knowing that you are the best there is!
Abstract—This paper will look at using open source tools (Blender © , LuxRender © , Open CV , SUMO , and Python © ) to build an image processing model for exploring combinations of sensors/platforms for any given image resolution. The model produces camera position, camera attitude, and synthetic camera data that can be used for exploitation purposes. We focus on electro-optical (EO) visible sensors to simplify the rendering but this work could be extended to use other rendering tools that support different modalities. Open Computer Vision (Open CV) is used to generate a camera model for the intrinsics of the virtual camera. This camera model is then used to geo-project the images from pixel space into a world coordinate system storing the output in National Imagery Transmission Format (NITF) for display with standard Intelligence Analysis (IA) tools such as RYA’s Pursuer . We also demonstrate Simulation of Urban Mobility (SUMO) software to simulate complex traffic patterns. The key idea of the paper is to provide an architecture for layered sensing simulation which is modular in design and constructed on open-source off-the-shelf software yielding a physics accurate virtual model of the world. This architecture shows how leveraging existing open-source software allows for practical layered sensing modeling to be rapidly assimilated and utilized in real-world applications. In this paper we demonstrate our model output is automatically exploitable by using generated data with an innovative geo- registration algorithm and real time display.
- Blender Camera Animation
- SUMO Animation
- Sadr City Car Animation Wire Frame
- Building Camera Animation
- Synthetic NITF In Pursuer Animation