Computer Vision Engineer

About the job

Company Description

At DefendEye, we design and manufacture rapid-deployment search and rescue drones. We build the tools that first responders rely on when every second counts. Our mission is simple: provide an autonomous "eye in the sky" to help locate and save lives in critical situations.

To do this, our drones need flawless visual perception—and that’s where you come in.

We are looking for a highly skilled Computer Vision Engineer to join our team in Krakow. Your very first mission? Develop the computer vision pipeline that allows our drones to autonomously identify, approach, and land perfectly inside a recovery box in outdoor environments.

🚀 What You’ll Do:

Vision Systems: Design and implement real-time computer vision algorithms from scratch using C, C++, and OpenCV.

Precision Navigation: Solve complex outdoor targeting problems, starting with teaching our drones to autonomously land in highly confined spaces.

Edge Computing: Deploy, test, and relentlessly optimize these models on resource-constrained edge hardware.

OS Integration: Build and maintain custom Yocto Linux distributions tailored specifically for our CV hardware.

✅ The "Must-Haves":

• Exceptional programming skills in C / C++ is a MUST

• Deep, advanced knowledge of Linux environments.

• Proven experience with Computer Vision (OpenCV) and vision-based targeting/navigation.

• Direct hands-on experience with either Robotics, Drones, or UAS (Unmanned Aerial Systems) platforms.

• Experience optimizing code for Edge Computing / Embedded Systems. (e.g., Raspberry Pi, custom embedded systems).

• Practical experience working with the Yocto Project.

• Experience with fiducial markers (AprilTags, ArUco) for precision landing.

⚠️ STRICT LOGISTICAL REQUIREMENTS:

Start Date: You MUST be able to start within the next 30 days.

Location: This is a full-time role based in Krakow, Poland. We offer a hybrid setup, but you MUST be in the office at least 3 days per week.

🚨 DO NOT APPLY if you cannot start within 30 days or cannot be in our Krakow office 3 days a week.

🌟 The "Nice-to-Haves":

Experience with ROS/ROS2, GStreamer, or integrating vision pipelines with flight controllers (ArduPilot/PX4).