II 2016 – II 2017        Photogrammetry Software

We wrote an innovative, high precision, 3D reconstruction software. Based on aerial images, using Structure from Motion algorithms, it reconstructs with 1 cm precision, the geometric structure of the photographed object.

XII 2015         LED bulletin board for students

We programmed a LED bulletin board that downloads information from a web page and displays it to students to encourage them to make their thesis in our department. It displays temperature, air pressure, time and some other text.

bulletinboard

The bulletin board is installed in one of Wroclaw University of Technology buildings.


II 2015 – VIII 2015         Autonomous Unmanned Aerial Vehicle

Our research team took part in a prestigious international contest https://valeoinnovationchallenge.valeo.com/ and worked on a very interesting project connected with aerial robotics and image processing. Over 1300 teams from the whole world participated in this contest, only 20 teams advanced to the semifinals, including us. In less than 15 weeks we ware able to build an unmanned aerial vehicle capable of avoiding obstacles using a LIDAR and write our complex software.

AUAV1

AUAV2


We prepared a film that shows the idea of this project

and a second one that shows the present technical level of our invention .

You can download the detailed description of this project from here: http://nice-idea.org/wp-content/uploads/2015/09/Nice-Idea.pdf . Unfortunately we run out of time and ware unable to implement interesting computer vision algorithms and that was probably the reason why we didn't advance to the final.


I 2015 - VI 2015 Obstacle avoidance algorithms using LIDAR data

To provide obstacle detection RPLidar 2D laser scanner was used to connect with ROS (Robot Operating System). The basic device handling the RPLidar can be found here. The code was extended by an additional node to provide a preview of the obstacle detected. RPlidar measurements are used to calculate the best avoidance direction when getting close to the obstacle. After detecting an object which crosses the safe zone, the algorithm calculates the direction in which the drone should move. The direction is calculated in such a way to allow the drone to stay at a safe distance and simultaneously reduce the amount of surplus route. The direction of the drone's movement is presented in the turtlesim node. We have prepared a short video which shows how the obstacle detection works. The second video presents the data output of this algorithm.


I 2015 - VI 2015 Visual environment recognition based on key points distribution

Experimental analysis was conducted for seven algorithms detecting keypoints: BRISK, FAST, GFTT, HARRIS, MSER, ORB and STAR. It was observed that the probability distributions of selected keypoints are drastically di erent for indoor and outdoor environments for all analyzed algorithms. This has lead to a universally e ective indoor/outdoor environment recognition method. It may prove to be a crucial step in the
design of robotic control algorithms based on video vision, especially for mobile robots.
indoorOutdoor
We ware able to distinguish environments using several different classifiers.
The effectiveness of the classification of indoor / outdoor reaches 94%.

classificationResults

Scientific paper : Keypoints for indoor and outdoor environments in video sequences


III 2015 – IX 2015        3D Printed Finger Prosthesis

We designed a finger prosthesis for a disabled man who lost his fingers in an accident. Several times we had to adjust and print upgraded versions of this prosthesis.

fingerPros

fingerPros2


III 2015 – VII 2015        3D Printed Hand Prosthesis

Together with the http://www.knezi.pwr.edu.pl/ we prepared a prototype of a low cost hand prosthesis equipped in muscle sensors and controlled by a ATMega168 microprocessor. You can watch here https://www.youtube.com/watch?v=QF0eoHqIKoI  how the prosthesis works.

handPros1

handPros2


I 2015 – II 2015        3D Printed CNC Router

To prepare complexly shaped elements our research team uses the K8200 3D printer and the Portal Cyclone CNC milling machine. Portal Cyclone is a 3D printed milling machine. We designed and built the Portal Cyclone ourselves by printing all its elements on K8200 3D printer.

CNCrouter1

CNCrouter2


You can check out how a Portal Cyclone milling machine works here https://www.youtube.com/watch?v=X5_7IKoARFU


XI  2014- XII  2014         Nice-Idea.org Web Page

To be able to present our projects to a wider audience we created a website http://nice-idea.org/ and a facebook fanpage https://www.facebook.com/scientificclubniceidea .


VI  2014 – XII 2014        Gender, Age and Emotion Recognition System

Our research team designed and implemented a gender, age and emotion recognition system intended to be installed in shops to analyze customer preferences and behavior. For feature recognition we used the OpenCV python library.

genderRecognition1

The gender, age and emotion recognition system works on an Odroid U3 embedded computer. It uses a Logitech C100 USB camera with modified optics.

genderRecognition3

genderRecognition2

You can download this app from https://github.com/bitcoinsoftware/Gender-Recognition


I  2014 – V 2014        Virtual Fitting Room

Our research team prepared a prototype of an app designed for people who would like to try on clothes without the need of walking into a store. The app finds automatically parts of the body and matches the clothing to the body.

vrFittingRoom1

vrFittingRoom2

vrFittingRoom3

You can download this app from https://github.com/bitcoinsoftware/virtualFitingRoom/


I 2013 – XII 2013        Betting Software

The software was prepared using python, sqlite3 and PyQt4. The aim of this app was to enable us to select soccer games in which there is an increased probability of a tie at the end of the match, analyze and simulate different betting strategies and predict the financial result of our gambling. You can download the app from  https://github.com/bitcoinsoftware/Soccer-Betting-Simulation . In this app you can set many different parameters that define your betting strategy. The picture shows how many values you can modify in your algorithm.

bettingSoftware1

Than you are able to simulate your betting strategy on past data and analyze the simulated cash flow.

bettingSoftware2
You can tell the program to find the most remunerative betting strategy and compare different strategies. Thanks to this program, we found out that probability of a draw is not the same throughout the entire football season. In the middle of the football season the number of draws is higher than in other periods. In the case of the second Italian division, the seasonal difference between the average probability of draw match result for the whole season and for the matches selected only from the middle of the football season is about 4 percentage points. Statistical analysis showed that there is a correlation between the score of the match, and the results obtained by the team in previous games. The correlation is associated with the condition of a sport team, which can be determined on the basis of several earlier matches. Studies have shown that there is increased probability of the draw result (for about two percentage points) if the result of the previous game played by the same team also was ended with a draw. We gambled using our algorithm and we won about 3000 Euro. That's the graph of our cash flow.
bettingSoftware3
Today we must say, that we ware lucky. That season, there ware a lot of draws in the French and Italian league. We wrote a scientific paper about that, you can download it here http://nice-idea.org/wp-content/uploads/2015/09/ko-oper_2015_Loziuk_v5.pdf


VIII 2012 - XII 2012        Interferometry Fringe Analysis Software

Our research team was responsible for preparation of an advanced scientific computer program intended to work with a Michelson interferometer build by Optel Opole company. Our software is used for conducting experiments in the Institute of Physics in Jan Kochanowski University .

fingerAnalyzis

You can download it from https://github.com/bitcoinsoftware/Fringe-analysis


I 2012 – VII 2012        Multidimensional Data Analysis and Visualization Software

Our research team prepared a sophisticated desktop app for multidimensional data visualization. We used several different visualization techniques such as isosurfaces, volumes, vector fields and plots.

visualization

You can download this software from here https://github.com/bitcoinsoftware/3D-Scientific-Visualization