Computer Vision Engineer
Computer Vision Engineer with focus on traditional computer vision techniques and algorithms. Candidate must have a thorough understanding computer vision fundamentals and hands-on implementation experience with a variety of traditional computer vision tasks. A strong conceptual understanding of multi-view geometry as it applies to computer vision is required. Candidate should also have enough knowledge of robotics to implement robotic perception algorithms that involve coordinate transforms in both 2D and 3D settings. Experience with robot operating systems and/or robotic hardware is not necessary. Candidate must be comfortable working in both C++ and Python environments.
A master’s degree with 3 years experience or equivalent required.
Practical implementation*** experience with some or all of the following is desired:
-syntactic feature extraction (ORB, FAST, etc.)
-image processing and manipulation (resizing, data type conversion, color space conversion)
-camera calibration
-image undistortion
-manipulating projection matrices
-working with quaternions
-conversion among and between coordinate systems
-homography computation
-image stitching
-bundle adjustment
-SLAM
-visual odometry
-structure from motion
***For all of the above mentioned skills, experience directly implementing an algorithm or through use of an existing library/toolset are both acceptable. This is a product development effort, not a research project.