Algorithm Engineer

OriginFlow; Hong Kong; On-site; HKD 600,000 - 900,000; Full time

Job summary

OriginFlow in Hong Kong seeks a robotics perception engineer to build end-to-end data pipelines and develop calibration and fusion algorithms. You will work across acquisition, processing, visualization, and validation stages to ensure robust perception systems.

The role emphasizes strong Python/C++ skills, OpenCV/PCL experience, and hands-on calibration with multi-sensor setups. Collaboration with hardware and software teams is expected in a fast-paced environment.

Qualifications

• Proficiency in Python and C++.
• Experience with OpenCV, PCL, NumPy, SciPy.
• Foundations in multi-view geometry, sensor fusion, and state estimation.
• Hands-on camera and sensor calibration experience.
• Experience with ToF, VIO, hand–eye calibration, or similar perception systems.
• Familiarity with ROS/ROS2 and robotic perception pipelines.

Responsibilities

• Build end-to-end data pipelines for acquisition, processing, visualization, and validation.
• Develop tools to quantify multi-sensor synchronization and alignment accuracy.
• Implement algorithms for fisheye calibration, image quality evaluation, and distortion correction.
• Develop ToF depth filtering, registration, and point cloud generation pipelines.
• Prototype multi-modal fusion approaches combining vision, depth, and IMU data.
• Design automated multi-sensor calibration systems and evaluate calibration quality.
• Produce feasibility analyses and performance reports to support product decisions.

Skills

Python

C++

OpenCV

PCL

NumPy

SciPy

Multi-view geometry

Sensor fusion

State estimation

Camera calibration

ToF

VIO

Hand–eye calibration

ROS/ROS2

Robotic perception

Tools

OpenCV

PCL

NumPy

SciPy

ROS2

Job description

The Missions

• Build end-to-end data pipelines for acquisition, processing, visualization, and validation.
• Develop tools to quantify multi-sensor synchronization and alignment accuracy.
• Implement algorithms for fisheye calibration, image quality evaluation, and distortion correction.
• Develop ToF depth filtering, registration, and point cloud generation pipelines.
• Prototype multi-modal fusion approaches combining vision, depth, and IMU data.
• Design automated multi-sensor calibration systems and evaluate calibration quality.
• Produce feasibility analyses and performance reports to support product decisions.

The DNA

• Strong proficiency in Python and C++.
• Experience with OpenCV, PCL, NumPy, SciPy, and related libraries.
• Solid foundations in multi-view geometry, sensor fusion, and state estimation.
• Hands-on experience with camera and sensor calibration.
• Experience with ToF, VIO, hand–eye calibration, or similar perception systems.
• Familiarity with ROS/ROS2 and robotic perception pipelines.