Full Stack Entwickler mit Schwerpunkt auf C# nicht verfügbar bis 30.11.2019

Full Stack Entwickler mit Schwerpunkt auf C#

nicht verfügbar bis 30.11.2019
Profilbild von Anonymes Profil, Full Stack Entwickler mit Schwerpunkt auf C#
  • 60439 Frankfurt am Main Freelancer in
  • Abschluss: M.Sc. Graphic Engineering and Design
  • Stunden-/Tagessatz: nicht angegeben
  • Sprachkenntnisse: deutsch (verhandlungssicher) | englisch (verhandlungssicher) | slovakisch (Muttersprache) | serbisch (Muttersprache)
  • Letztes Update: 03.05.2019
SCHLAGWORTE
PROFILBILD
Profilbild von Anonymes Profil, Full Stack Entwickler mit Schwerpunkt auf C#
DATEIANLAGEN
Profil 1008 - Full Stack / C# Entwickler (.docx)
Der Download dieser Datei ist nur mit einem kostenpflichtigen Account möglich.
Profil 1008 - Full Stack / C# Entwickler (.pdf)
Der Download dieser Datei ist nur mit einem kostenpflichtigen Account möglich.
SKILLS
  • C#
  • Multithreading
  • Bootstrap.js
  • ASP.NET
  • Electron.NET
  • jQuery.js
  • Windows Forms App (.NET)
  • MS Unit Test
  • MSSQL
  • ASP.NET Web Forms
  • EmguCV.NET
  • Techan.js
  • ASP.NET MVC
  • Accord.NET
  • WebSocket
  • JSON (Newtonsoft Json)
  • AForge.NET
  • LittleCMS
  • LINQ C#
  • JavaScript
  • Babylon.js
  • Entity Framework
  • HTML5/CSS3
  • Microsoft Git
  • Rest API
  • D3.js
  • OWIN OAuth
  • Visual Studio Installer Projects
  • Graphosaurus.js (WebGL)
  • Source Tree
  • Swift
  • Objective-C
  • Scrum
  • Matlab/C++
  • WPF.NET
     
  • Visual Studio
  • SQL Server Management Studio
  • .NET Framework
  • .NET Core
  • .NET Standard
  • Office package
  • Adobe Illustrator
  • Adobe Photoshop
  • Adobe InDesign
  • XCode
  • AutoCAD
  • GIMP2
REFERENZEN
  1. Color and Appearance Map, from 09/2015 till 03/2016, from 10/2016 till 03/2019
  • C#, JavaScript, HTML/CSS, MSSQL, ASP.NET Web Forms
  • AForge.NET, bootstrap.js, D3.js, Entity Framework, EPPlus.NET, jQuery.js, NewtonsoftJson.NET
The main feature of this solution is to extract and cluster colors with machine learning algorithm. Project is used to show potential customers, which data can be extracted from simple image, and how further running this data through customized algorithm can show hidden data which normally are not evident. Developed were algorithms specialized for clustering hair, skin colors, and also algorithm for general use. The algorithm for hair interestingly enough detected harmonic distribution of Melanin in Hair which potentially could be attributed to albinism.
 
  1. Color and Appearance Map, custom version, from 02/2016 till 04/2016
  • C#, JavaScript, HTML/CSS, JSON, ASP.NET MVC
  • bootstrap.js, D3.js, Entity Framework, jQuery.js, NewtonsoftJson.NET
This application is used to show results of the project 1. above to potential customers. Special algorithm is written in Matlab, through which all the necessary images are processed. The yielding results of the algorithm are saved as .txt in JSON format. A web site as MVC project was created, importing JSON file as source for the frontend to display the results. A complex frontend was developed in order for a special chart created by using D3.js library to display the algorithm results. A merging image algorithm for simulation is created to show the results in real-life environment.
 
  1. 3M ESPE ToothShade Catcher, from 02/2016 till 3/2016, from 08/2016 till 9/2016
  • Swift, Matlab/C
Application is taking photo of tooth with characterization stick and is transforming the photo to a single dimension Uint8 array, which is used as input vector to C code. The challenge was creating a vector with right order of integers for C code to digest the input vector and also fetching the output vector and transforming it back to usable photo for user to see. Conducted also by importing of Matlab/C algorithm into Swift.
 
  1. Roche Accu-Check SmartScan, from 03/2016 till 04/2016
  • ObjectiveC, Matlab/C
Objective-C Version, similar to the project 3. above. Created as a mockup for client to see the design, and possible functionality and to determine whether a connection between Objective-C and C is more feasible than the connection between C and Swift. Objective-C won.
 
  1. Cosnova, from 04/2016 till 06/2016
  • C#, C++, JavaScript, HTML/CSS, ASP.NET Web Forms
  • bootstrap.js, jQuery.js, pure.js, DocumentFormat.OpenXml.NET, LittleCMS, SharpZipLib.NET
Complex project requiring advance knowledge of mathematic, due to complex transformation between ICC profiles. The goal is to allow a user to enter Lab values of any color imaginable and to get result in sRGB and other standardized color spaces, as well as in hexadecimal representation of a particular color. For the transformation is used C++ LittleCMS library for which the wrapper is written in C#, however all transformations are recalculated manually to confirm the accuracy of LittleCMS library. A user is able to download the Excel file with the results. Connecting C# with C++ was especially challenging task.
 
  1. Image Profiler, from 06/2016 till 09/2016 
  • C#, JavaScript, HTML/CSS, WebFroms
  • jQuery.js, D3, OWIN OAuth 2.0, babylon.js, graphosaurus.js
This project is next step in evolution of Color and Appearance Map project. The main goal is to not only represent few extracted colors but also the value of each pixels represented in 2D graphs and also representing whole 2D image in 3D color space. Everything is written from a scratch. A security protocol (OWIN) is implemented for a login. The image pixels are read and cleaned with noise removing algorithm, then they are sent to frontend in array to be used by 3D.js and Graphosaurus.js to get the desired output.
 
  1. SCiO project, from 09/2016 till 10/2016
  • ObjectiveC
  • SCiO SDK
The SCiO product Molecular Sensor is a product the manufacturer claims it measures molecular structure of any material. The goal is to reverse engineer it and look what makes it ticks. Manufacturer published the SDK to allow for advance users to import their algorithm for data clustering and data processing. After testing it for a while, at the end, a discovery was made that the measurement is made with simple CMOS sensor which has filter that blocks visible wavelength and only IR radiation falls on to the sensor. This discovery was a base for further researches in this field.
  1. Color Catcher 2.0, from 01/2017 till 02/2017
  • ObjectiveC/Swift
The ColorCatcher is product themed in candidate’s master thesis. The end product can be obtained under:
https://www.techkon.com/colorcatcher-106.html
The preparations are made for implementation of new algorithms into Color Catcher and the behavior of new algorithms is tested.
  1. DSL Software, from 01/2017 till 06/2017
  • C#, C++, Windows Forms.
  • EmguCV, ZedGraph, Multithreading, Vimba.NET, Visual Studio Installer, LittleCMS
The DSL Software is candidate’s first desktop application written in Visual Studio. The goal of the project is to characterize the industrial Mako camera, so that the colors from captured frames can be extracted independently of illumination. This is archived using characterization card and manually setting the red, blue sensor sensitivity as well as exposure time of camera. From the frames the characterization card is automatically detected, which is achieved using code written in C++ and compiled to .exe. Furthermore this .exe is connected with C# so, that the .exe is fired up. The process is fully automated, the four frames are extracted, characterized and transformed into an ICC profile to compensate the illumination. Also, the spectrophotometer is connected with the code so that measurement is automatized also through the C# code. Both measurements, the images results and spectrophotometers results, are sent to a machine learning algorithm to do process it, hereafter results are saved in a folder.
 
  1. Color Finder, from 01/2017 till 10/2017
  • C#, C++, Windows Forms.
  • EmguCV, ZedGraph, Vimba.NET, Multithreading, Visual Studio Installer
Color Finder is used as a presentation project. The user is able to load its image to the software and extract the region of interest (ROI) by clicking on user interface, which contains uploaded image. The ROI is circular in shape, which indicates to the user, which of the pixels are accounted for the measurement. Values in the ROI are averaged and single average color is presented in CIE Lab color space. User is further able to choose, from available color spaces, to which specific color space the averaged color should be transformed. Before any measurement takes place, user is also able to import its own color space to the software, for the software to use it as a starting color space for transformations. Explained on an example of a color image, which is scanned and reprinted. In practice these two images are never the same due to false color transformations between color spaces. But, by using color space specifically created for scanner and for printer respectively, the reprinted image can be identical as the original one. This software can also allow a user to see beforehand which color will be reprinted.
 
  1. C&A Map Custom (see project 2), from 02/2017 till 03/2017
  • C#, JavaScript, HTML/CSS, MSSQL, WebForms.
  • AForge, Ajax Control Toolkit, Bootstrap, D3, EntityFramework, EPPlus, jQuery, NewtonsoftJson.
Original C&A Map code is forked and rewritten so, that more images at same time can be processed. An additional algorithm is built to handle texture of leather and artificial leather. The goal is to distinguish artificial leather from real one in automobile industry.


13. Oracle Database DLL, from 03/2017 till 05/2017
  • C#
  • Oracle
A Module for Oracle database connection is written. Whilst the structure of database is not complex, a complex data is required to be written in a database structure. This is archived by saving JSON formatted text to a string column.
  1. K-means algorithm, from 04/2017 till 05/2017
  • C#
  • MSUnit Testing, EmguCV
The algorithm is written as a separate module to be used in other projects. K-means is used in data clustering. The reason for using this algorithm in color clustering and in data clustering in general is the speed of processing which can be easily adjusted, by using another algorithm to pre-process data to reduce iteration in K-means itself.
  1. SKZ Clustering, from 06/2017 till 02/2018
  • C#, Windows Forms App (.NET)
  • EmguCV.NET, AForge.NET, EPPlus.NET, LittleCMS, Multithreading, Visual Studio Installer
In this project main requirement is to allow a user to cluster a profiled image with several different algorithms. It is needed for a program to find ratio of colors on the image, used for images of shredded plastic for recycling. The goal is to make separation of garbage scraps by color with a first step being a determination of the amount of each color on recycling conveyer line.
  1.  AVON DSL Software, from 06/2017 till 03/2018
  • C#, C/C++, Windows Forms App (.NET), Arduino, Oracle
  • Accord.NET, EmguCV.NET, EPPlus.NET, Newtonsoft.Json.NET, LittleCMS, Vimba.NET, Core.Animation.NET, Multithreading, Visual Studio Installer
Candidate`s most complex project. The goal is to measure the color, gloss and glitter of certain products. This is archived using lasers and illumination at specific angle. Five images in a row are taken automatically, each image is taken under different illumination and with different lasers, which are controlled with Arduino electronic switcher which is connected through a C# DLL module. After all images are taken, the algorithms clean the images, extract the ROI, calibrate and characterize the image and display results to a user. A special challenge is creating of hidden settings for a recalibration of the system during the test measurement. The number of variables in this project is overwhelming.
  1. ZIM Project, from 10/2017 till 11/2017
  • C#, C++, Windows Forms App (.NET)
  • Accord.NET, EmguCV.NET, LittleCMS, Multithreading, Visual Studio Installer
An algorithm for clustering of colors is written with background extraction so, that the color of background is not accounted during colors extraction. After extraction of ROI and characterization of image, the background image is sampled in 4 different placed outside of ROI and these values are used as parameter to algorithm for it to ignore background colors. The user has several features at a disposal helping him to decide how much clusters and which algorithm are most optimal.
  1. Color Transformations, from 11/2017 till 12/2017
  • C#
  • MSUnit Testing
The DLL is created from transformation formulas for LAB, RGB, XYZ color spaces with specific illumination D65 and standard observer of 2 angles and 10 angles respectively. The goal is to encapsulate the complex formulas in one DLL and to import them easily in future projects. Also, it is planned that module encapsulates ICC profile transformations in C# for future uses, since this kind of DLLs is very expensive.
 
  1. Color Finder, 2.0 from 05/2018 till 06/2018
  • C#, C++, Windows Forms App (.NET)
  • EmguCV.NET, Multithreading, Visual Studio Installer
Practical application of first Color Finder application. The goal is to extract information in defined circle of a pointed position in an image, where the circle size is adjustable. After the ROI extraction the user gets the averaged color of the ROI and the color is presented in LAB, RGB and ISO color spaces. The user can choose to which color space the color is to be transformed, because the amount of color space is vast. Additionally, a user can save all the results.
 
  1. Crypto Trading Bot (a private project), from 02/2018, ongoing
  • C#, REST API, WebSocket, Core 2.0/Core 2.1, Electron
  • EntityFrameworks, Newtonsoft.Json, Visual Studio Installer
Candidate’s personal hobby project. The goal is a trading algorithm embeded in the software to automatically trade cryptocurrencies on exchange markets. For it to communicate with markets, the REST Api is used. At first, the prices were streamed via WebSocket, but the socket seemed not to be reliable. In the second phase updates of prices are migrated to a ticker. Ticker: background prices are checked every x-number of seconds through a REST Api. To show the results on the chart a Windows app is created, digesting JavaScript code through Electron Library. After Core.NET turned out to be unreliable and not as flexible as Framework.NET 4.7.2, JavaScript code is abandoned, and the frontend is written in C#.
KONTAKTANFRAGE VERSENDEN

Ja, ich akzeptiere die AGB

Ja, ich akzeptiere die Datenschutzbestimmungen

Hinweis: der Versand ihrer Kontaktanfrage ist komplett kostenfrei