Welcome to Tramacsoft,
your future software provider.

SERVICES

We provide a wide range of services for Vision and Robotic applications:

CUSTOM SOFTWARE DEVELOPMENT

 

We design and develop software for both general-propose and task-specific vision and robotics applications, working with state-of-the-art technology to provide the most efficient and reliable solutions.

The services in software development include analysis, planning, design, implementation, integration, testing, maintenance and upgrading.

CONSULTING

 

Based on our experience and technical background, we provide external consulting for a wide range of machine vision and artificial intelligence applications.

Feel free to contact us for an early assessment and discussing the most optimal strategy to tackle your problems without any commitment.

 Contact us!

OPTIMIZATION

 

We provide an optimization service to increase the efficiency of existing machine vision solutions, improving robustness, accuracy and computational limitations.

Our team of experts can investigate new hardware setups to reach your desired targets or add-ons to further optimize your current solutions.

About

Tramacsoft is a specialized Vision and Robotics company, located in Frankfurt am Main, Germany, that offers highly technical services for a wide range problems. We provide services ranging from software development and external consulting to fine tuning of systems to achieve the most optimal results. We focus on providing high quality services based on state-of-the-art algorithms and high-end technology applied to practical scenarios, defining reliable solutions that maximize efficiency, adaptability and robustness to overcome existing productivity limitations.

We leverage our experience working in very diverse industry and cutting edge R&D to help you with your challenges.

Computer Vision / Image Processing
Performance optimization
3D reconstruction
Machine Vision
Machine Learning / Deep Learning
Odometry / SLAM / Localization & Navigation

Computer Vision and Machine Vision

We apply the most robust computer vision algorithms to solve practical problem for industrial scenarios. We have experience developing all type of image processing solutions as independent products or add-ons for a wide range of commercial solutions (Cognex, Halcon, NI Vision, etc) and open-source frameworks (ROS, ROS-Industrial).

 

Machine Learning and Deep Learning

We employ novel machine learning approaches to develop adaptable solutions that fit our customers requirements. We developed a rendering pipeline that allow to highly automatize the most labour intensive tasks reducing the main bottleneck of Deep Learning industrial expansion.

 

User-friendly solutions

We prioritize understanding the final target needs and requirements to design and develop user-friendly solutions and provide insightful, practical and relevant consulting services. Our user-friendly developing services includes integrated cloud data storing, structuring and managing, and web or mobile interfaces.

Portfolio

vaccum_cleaning_robot

Vacuum cleaning robot

Assist in the development of mapping and navigation software

SLAM Mobile robots C++ Docker Unittest CI/CD ARM

object_pose_estimation

Bin picking

Object localization for challenging scenarios with multiple cameras

Docker Rendering Python Unittest C++ CI/CD Pybind

Redesign automotive software module

Refactor large data processing module while preserving precision and speed

C++ ADAS Clean code Profiling Visual Studio Autonomous Driving

Tool to process logs

Facilitate debugging by creating a tool to post-process logs and replicate edge scenarios

ROS C++ OpenCV Logs

legoChecker

Brickinspector (More info)

Develop application to segment and classify Lego ® pieces to verify completness of the product sets

Python PyTorch Semantic Segmentation QT NextCloud Windows Linux Deep Learning Machine Learning

3D Laser Scanning

Industrial laser 3D scanner

Support the development of laser processing library to generate 3D pointclouds by optimizing speed and code mantainance

Matlab Coder C++ VTune profiler Matlab Python Unittest CI/CD

photorealistic_rendering

Photo-realistic rendering for ML

Rendering pipeline to generate ground truth and drastically reduce the need of manually annotated data for Deep Learning applications

Python Deep Learning PyTorch Machine Learning Artificial data

image_editor

Image editor

Develop image edition application to solve Augmented Reality limitations

Python OpenCV SIFT Descriptors QT PyInstaller

localization_particle_filter

User Localization

Implement algorithm to localize user in a mapped environment

C# Unity Localization Particle Filter Navigation

Tracking of body motion

Scan user skeleton in real time to create diverse interactive booths displaying digital avatars in 2D or creating the illusion of a 3D object

Processing OpenCV OpenNI Kinect Asus Xtion 3D model Eye tracking

Gts_robot_tracking

GTS: visual tracking of robots

Contribute to open source project automatizing tracking, HW support and bug fixing

C++ Linux Windows Cmake Visual Studio Github Ethernet cameras Calibration

restoration_colorize

Digitalization and restoration of videos

Digitalize VHS and Video-8 tapes and enhance them using Deep Learning restoration techniques

VHS Video8 Deoldify Deep Learning Video processing

IMU mechanical validation

Perform mechanical validation of Inertial Measurement Unit mounted on CNC

G-Code Python Serial Bus Reverse engineering IMU Gyroscope

ground_plane_estimation

Automotive SLAM

Research algorithmic improvements for Pose estimation and 3D reconstructions

C++ Automotive 3D reconstruction SLAM Visual Odometry Optical Flow OpenCV

Code maintenance

Code refactoring, static analysis, profiling and algorithmic optimization to avoid hardware limitations

C++ Python Git Cmake Profiling

elderly_monitoring

Monitoring seniors living alone

Create behaviour reports from a network of sensors to detect accidents and health issues

Linux Nagios MySql Python HTML

pybind_py_cpp

Binding C/C++ code to Python

Implementation of a Python wrapper of a C/C++ library for a 100k+ codebase

C++ C Python Pybind11 CMake Jenkins CI/CD GoogleTests

3d_scanning_iphone

Phone based 3D scanning

Improve accuracy, speed, server integration, code cleanliness of pipeline for scanning of objects in 3D

C++ OpenCV Point Cloud Library (PCL) Postman Valgrind Profiling iPhone 3D scanning Stitching

relocalization_odometry_alignment

Re-localization

Estimate current vehicle position in a previous scanned environment

C++ GPS Protocol Buffer Serialization Automotive Odometry

Contact

Tramacsoft GmbH

Schloßstraße 52
60486 Frankfurt am Main
Germany

 

Amtsgericht Frankfurt am Main
HRB 114873
VAT ID DE 324 815 202
Managing Director: Marc Barnada Rius

 

Email: info@tramacsoft.com