# Object tracking system **URL:** \ **Category:** PX4 Autopilot\ **Created:** [April 23, 2021, 5:55am UTC](https://discuss.px4.io/t/object-tracking-system/22083 "2021-04-23T05:55:55Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![dark](https://discuss.px4.io/letter_avatar_proxy/v4/letter/d/a88e4f/32.png) [@dark](https://discuss.px4.io/u/dark)\ **Post date:** [April 23, 2021, 5:55am UTC](https://discuss.px4.io/t/object-tracking-system/22083/1 "2021-04-23T05:55:55Z")
Hello I am working on my university graduation project, which is as follows: I am building a plane with a px4, I also connected it to a jetson nano kit And I connected it to a camera with an object tracking system my question is: Can I control the drone through the (computer vision) (object tracking) in the camera? Or should I start programming a new object tracking? thanks for the help ---
**Author:** ![rajeshroy402](https://discuss.px4.io/user_avatar/discuss.px4.io/rajeshroy402/32/8453_2.png) [@rajeshroy402](https://discuss.px4.io/u/rajeshroy402)\ **Post date:** [April 23, 2021, 11:59am UTC](https://discuss.px4.io/t/object-tracking-system/22083/2 "2021-04-23T11:59:41Z")
Hi Dark, do you have any follow up to this thread? ---
**Author:** ![dark](https://discuss.px4.io/letter_avatar_proxy/v4/letter/d/a88e4f/32.png) [@dark](https://discuss.px4.io/u/dark)\ **Post date:** [April 23, 2021, 12:48pm UTC](https://discuss.px4.io/t/object-tracking-system/22083/3 "2021-04-23T12:48:49Z")
Just starting, and I want help to know what the main things I have to focus on ---
**Author:** ![rajeshroy402](https://discuss.px4.io/user_avatar/discuss.px4.io/rajeshroy402/32/8453_2.png) [@rajeshroy402](https://discuss.px4.io/u/rajeshroy402)\ **Post date:** [April 23, 2021, 1:00pm UTC](https://discuss.px4.io/t/object-tracking-system/22083/4 "2021-04-23T13:00:59Z")
> [@dark](#): > > Can I control the drone through the (computer vision) (object tracking) in the camera? Yes you can use Pixhawk FC and Nano for this task! May I know more about the model you are using and whether it is just the tracking as your end objective by the camera feed or something else? ---
**Author:** ![Tolgahan\_SARP](https://discuss.px4.io/user_avatar/discuss.px4.io/tolgahan_sarp/32/21482_2.png) [@Tolgahan\_SARP](https://discuss.px4.io/u/Tolgahan_SARP)\ **Post date:** [April 22, 2026, 6:25am UTC](https://discuss.px4.io/t/object-tracking-system/22083/5 "2026-04-22T06:25:19Z")
Hi everyone, I’m working on a UAV project where I want to **detect, track, and follow a moving object without using GPS** (fully GPS-denied environment). I would really appreciate guidance on both **camera selection** and **localization strategy**. * * * ### 🧩 Current System Architecture - **Flight Controller:** Pixhawk running PX4 - **Companion Computer:** Considering * * * ### 📷 Camera Options I’m Considering **Option 1: Global Shutter Monocular Camera** - Pros: No motion distortion, simpler pipeline - Cons: No direct depth information **Option 2: Luxonis OAK-D Lite (Stereo + RGB)** - Has stereo depth + onboard AI acceleration - Includes global shutter mono sensors for depth - Pros: Depth estimation + object detection on-device - Cons: More complex integration, potential latency * * * ### 🎯 Goal - Detect and track a target (person/vehicle/etc.) - Estimate relative position - Control the drone to **follow the target autonomously** - Operate fully **without GPS** * * * ### ❗ Main Challenge: Localization in GPS-Denied Environment This is where I’m struggling. If I use: - **Monocular camera:** I only get 2D tracking (no scale/depth) - **Stereo camera (OAK-D Lite):** I get depth, but I’m unsure how reliable it is during motion and at longer distances * * * ### ❓ Questions 1. **Camera Choice:** Would you recommend a **global shutter monocular camera + VIO** or a **stereo system like OAK-D Lite** for this use case? 2. **Localization Approach:** In a GPS-denied setup, what is the most robust way to estimate position for object following? 3. **Tracking + Control:** What is the best approach to convert: 4. **Practical Experience:** Has anyone built a similar system? Any issues with: * * * ### 💡 Additional Notes - Real-time performance is important - The drone will operate in **dynamic environments** - I can design the rest of the hardware (power, mounting, etc.) * * * ---
**Author:** ![Jaime\_Garcia](https://discuss.px4.io/user_avatar/discuss.px4.io/jaime_garcia/32/22070_2.png) [@Jaime\_Garcia](https://discuss.px4.io/u/Jaime_Garcia)\ **Post date:** [September 21, 2026, 10:51am UTC](https://discuss.px4.io/t/object-tracking-system/22083/6 "2026-09-21T10:51:31Z")
Hola, has conseguido algo, yo estoy jhhacxiendo lo mismo para un proyecto final de carrera. ---
**Author:** ![yigitcan-ozturk](https://discuss.px4.io/user_avatar/discuss.px4.io/yigitcan-ozturk/32/22039_2.png) [@yigitcan-ozturk](https://discuss.px4.io/u/yigitcan-ozturk)\ **Post date:** [October 19, 2026, 5:35am UTC](https://discuss.px4.io/t/object-tracking-system/22083/7 "2026-10-19T05:35:11Z")
Hi @Tolgahan_SARP, For a GPS-denied object-following project, I would separate the problem into three parts: vehicle localization, target-relative estimation, and flight control. A stereo camera such as the OAK-D Lite can provide useful depth information, but its effective range and reliability depend heavily on lighting, texture, target distance, and motion. A monocular global-shutter camera combined with VIO can be a reasonable localization approach, although monocular scale observability and drift need careful consideration. My suggestion would be to start with: 1. Establish stable vehicle state estimation using visual-inertial odometry, independently of the target detector. 2. Estimate the target’s relative position and velocity, including uncertainty and measurement timestamps. 3. Validate the tracking pipeline in simulation before connecting it to PX4 Offboard control. 4. Explicitly handle stale measurements, tracking loss, and localization degradation. A valid target detection alone should not be treated as sufficient for autonomous following. One important distinction: target-relative tracking does not automatically provide a reliable vehicle position estimate in a GPS-denied environment. Before selecting the camera, could you clarify the expected target distance, vehicle speed, and whether the operation is indoors or outdoors? These constraints will strongly influence whether stereo depth, VIO, or another localization approach is appropriate. Best regards, Yiğitcan Öztürk PAMILANGA LTD