Sensor fusion algorithm for different types of sensors (radars, lidars, cameras)
Contribution:
Radar and camera fusion component for reducing the distance error (camera error)
R&D – composing matrices for different types of measurements,
C++ implementation with ROS communication and visualization part for displaying fused data.
C++/Cuda implementation. Bazel used for integration. Developed unit-tests
Preliminary check system for the sensors (pre-drive sensors check) R&D, C++ implementation with ROS communication, PyQT UI. Bazel used for
integration Investigation of the non-deterministic behavior in the complex objects tracking component
Debugging with gdb, inspecting the algorithms, analyzing the ROS communication delays expenses and code
architecture. Providing the recommendation how to fix the problem. Regression test tool for object tracking
module with testing episode to check the tracking and object recognition algorithms Tool developed in python
and is able to control and observe object tracking processes, records data results using ROS. Developed test
suites mechanism that provides an ability to write tests to check for the non-regression. In addition UI developed
in Qt DBScan optimized algorithm for image clustering using CUDA (small experience) Brainstorming ideas on
how to optimize the algorithm Tool for codebase module dependencies scanning Implemented using python and
bazel query engine. Used for recognizing transitive and direct dependencies. Provides user-friendly summary
about dependencies in the codebase. Unit-tests and refactoring for different modules in object tracking
component and geometry estimator Understanding algorithms. Functional and acceptance unit-tests. Clean code
refactoring
Technologies:
C++11, c++17, Sensor/data fusion , Debugging, Model based estimation , Predictive analytics, Camera, Optimization,
CUDA, GitHub, C++, Development of Lidar Algorithms, Boost, Software Architecture, LiDAR, Linux, CodeBeamer, Radar,
Python, Camera sensors,ROS, R&D , Conf1uence,Python Qt,Embedded, C++14
Involvement:
11/03/2019 — 24/04/2020