Vision-Based Tracking Turret using ESP32-CAM and YOLO

In our project, a low-cost wireless pan-tilt turret was designed and tested using an ESP32-CAM for video streaming and a host PC running YOLO for detection. Unlike monolithic embedded trackers, the decoupled architecture keeps heavy neural-network inference away from the motor-control path, so the microcontroller only streams video and drives servos. The system uses off-the-shelf parts costing about INR 1,940, and it can detect semantic object classes instead of just moving pixels. It achieved detection confidence above 0.85 and detection rates above 93% up to 2.5 m in well-lit rooms, with an end-to-end latency of about 80–150 ms. Performance degrades in low light and at long range because of the resolution and sensitivity limits of the OV2640 sensor. Overall, the proposed turret is a flexible and affordable platform for surveillance, human-robot interaction and academic research. In the future, we can add a Kalman filter [11] or a learned motion predictor to bridge missed detections, use a thermal sensor such as MLX90640 for dark environments, track multiple targets with a priority rule, and move inference to an edge accelerator such as Jetson Nano or Coral to remove the PC and Wi-Fi delay. A web dashboard for remote monitoring and manual override would further improve usability.

1.      Viola, Paul, and Michael J. Jones. “Robust real-time face detection.” International Journal of Computer Vision 57.2 (2004): 137–154. 2.      Dalal, Navneet, and Bill Triggs. “Histograms of oriented gradients for human detection.” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Vol. 1. 2005. 3.      Liu, Wei, et al. “SSD: Single shot multibox detector.” European Conference on Computer Vision (ECCV). Springer, 2016. 4.      Redmon, Joseph, et al. “You only look once: Unified, real-time object detection.” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2016. 5.      Ultralytics. “YOLOv8 Documentation.” docs.ultralytics.com. Accessed 2024. 6.      Bradski, Gary, and Adrian Kaehler. Learning OpenCV: Computer Vision with the OpenCV Library. O’Reilly Media, 2008. 7.      Espressif Systems. ESP32 Technical Reference Manual, Version 5.1. Espressif Systems, Shanghai, 2023. 8.      OmniVision Technologies. OV2640 Camera Chip Datasheet. OmniVision Technologies Inc., 2006. 9.      PySerial Development Team. “PySerial Documentation.” pyserial.readthedocs.io. Accessed 2024. 10.   Arduino. “ESP32Servo Library Documentation.” madhephaestus.github.io/ESP32Servo. Accessed 2024.