Francisco Lamosa

Develop plan for integrating sensors for navigation of tractors in closed fields: stereo vision, GNSS, radar, lidar, ultrasonics; develop plan for the use of aerial drone data with GNSS and mapping data for route planning and navigation Develop test plans for acceptance testing for system integration Develop multi-sensor calibration protocols and validate

Augmented Reality, Virtual Reality - Computer vision algorithm analysis and development

Contract through March 2017 - with option to change with 30 days notice • Investigated noise reduction algorithms (with Matlab, C++, CUDA) • Implementation of noise filtering in CUDA • Assumed responsibility for maintenance of visualization module (OpenGL-based DLL and C# for GUI) • Investigated 3-D segmentation of point cloud for registration of point cloud to CAD model

Investigating new image analysis algorithms, designing software for new modules, analysing client needs to provide solutions that meet client's needs. Working with partners and clients to define requirements and specifications, deliver custom solutions, and provide the training on the tool/SDK.

Developed tools for tracing OpenCL API calls on both the GPU and CPU – one significant feature was a measure of GPU processing element utilization (occupancy). Implemented unit tests for these features using the Google testing framework Investigated mocking tools Implemented testing/validation framework for hardware debugging features, such as ISA patching Writing trap handlers using the GPU ISA for hardware debugger

Ported Matlab code into C++ Developed an in-house C++ image processing library to add functionality missing in the default CImg class template library-both class library architecture and C interface Developed guidelines and standards for software development Evaluated different tools to assist in the development process: UML tools, prototyping environments (Scilab, Python, Octave), image processing libraries and packages Carried out tests improving stereo accura

Developed image processing algorithms on NVIDIA GPU using CUDA development environment (both runtime and driver API) for unsharp masking, contrast enhancement, and enhancing image stabilization algorithm; prototyping done in Matlab Integrated these algorithms into C++ based software libraries (with a C API) Wrote application notes and documentation for the new functionality

Developed image processing algorithms in C for varied image processing applications using Matlab and Visual Studio; also defined the multithreaded architecture for the different tasks of the traffic analysis system: capture, analysis, and data output Added support functions for HMCE/HMCR number plate recognition software: DES encryption, TCP/IP communication

Carried out evaluations of computer vision/image processing problems and provided existing/potential customers with solutions by addressing technical questions and providing code (in MFC or ANSI C/C++); carried out onsite training/support implemented system performance features required by customers Developed interfaces between video cameras and frame grabbers (frame grabber configuration files) handled customer issues by validating reported problems and coordinating the different development teams

Developed machine vision applications for customers under Windows (MFC-based); delivered quality control systems in the automotive(gauge calibration, print quality) and military sectors (assembly testing), automated research test platform video data analysis Developed user requirements specification, systems integration with external components, acceptance testing, and customer training Provided on-site/in-house tech support/training

Fusion of Multiple Camera Data for Traffic Monitorin Multiple camera images were registered by re-projecting onto a common reference plane using planar projective transformations to compensate for missing information in single images. Non-planar distortions were compensated by using multiple planes