Droneer AI VisionCube AI Perception Module for FPV Drones - Visible/ Dual Visible/ Thermal (ST/DT) Options

OverviewDroneer AI VisionCube is an AI perception module for FPV drones, available in visible-light (single/dual) and visible + thermal configurations. Depending on the selected model, it provides onboard recognition, dynamic tracking, and target lock functions for aerial applications, with on-screen display (OSD) support.Key Features Lightweight all-in-one integration; modular design with total weight <200g Compatible with BetaFlight/ArduPilot and other open-source flight control systems High-definition camera with ultra-low latency video output (<60ms); built-in Picture-in-Picture (PiP) mode; end-to-end latency below 60ms Core functions: 1.2km ultra-long range recognition, multi-target recognition, in-frame Picture-in-Picture, adaptive tracking, dynamic memory lock-on, multi-spectrum compatibility In-frame tracking architecture: full-screen landscape view with an 8x electronic zoom inset Multi-target recognition: powered by YOLOv7 depth-optimized multi-target detection; supports recognition of up to 50 dynamic targets and deployment of self-training algorithms Adaptive suction attachment tracking: locks onto the target in the center of the frame in as little as 0.5 second; target-locking success rate in high-speed, dynamic scenarios to 92% Dynamic memory lock-on: motion prediction + Kalman filter; 3-second re-capture upon target occlusion/leaving the field of view; tracking stability improved by 40% Multi-spectral sensing (thermal variants): visible light / infrared dual-mode sensor switching; uncooled infrared thermal imager; operating range -20 C to 60 C Model Options Variant Visible Imaging Thermal Imaging AI VisionCube S Single visible lens No AI VisionCube D Dual visible lenses No AI VisionCube ST Single visible lens Yes (thermal) AI VisionCube DT Dual visible lenses Yes (thermal) AI ST Pro Single visible lens Yes ("640" thermal, per variant name) AI DT Pro Single visible lens (per variant name) Yes ("640" thermal, per variant name) SpecificationsRecognition & Tracking (AI VisionCube S / D / ST / DT) Parameter AI VisionCube S AI VisionCube D AI VisionCube ST AI VisionCube DT Supported detection target recognition types Default: Vehicles Default: Vehicles Default: Vehicles Default: Vehicles Supported target recognition distance Vehicle: 450 m; Person: 170 m Vehicle: 1200 m; Person: 500 m Vehicle: 450 m; Person: 170 m Vehicle: 1200 m; Person: 500 m Minimum target recognition pixel 16x16 pixels 16x16 pixels 16x16 pixels 16x16 pixels Maximum dynamic tracking speed 60 km/h 60 km/h 60 km/h 60 km/h Intelligent suction tracking Supports crosshair auto-lock, enabling precise star-distance tracking and suction Supports crosshair auto-lock, enabling precise star-distance tracking and suction Supports crosshair auto-lock, enabling precise star-distance tracking and suction Supports crosshair auto-lock, enabling precise star-distance tracking and suction Trajectory prediction & memory tracking Supports trajectory prediction and feature memory dual-mechanism tracking Supports trajectory prediction and feature memory dual-mechanism tracking Supports trajectory prediction and feature memory dual-mechanism tracking Supports trajectory prediction and feature memory dual-mechanism tracking Maximum number of supported targets Up to 50 targets Up to 50 targets Up to 50 targets Up to 50 targets On-screen display (OSD) Supported Supported Supported Supported AI Image Processing Board Parameter AI VisionCube S AI VisionCube D AI VisionCube ST AI VisionCube DT Computing power (TOPS) 1 6 1 6 Control protocol CRSF CRSF CRSF CRSF Supported firmware BetaFlight / ArduPilot BetaFlight / ArduPilot BetaFlight / ArduPilot BetaFlight / ArduPilot Input voltage (V) 9~16 9~16 9~16 9~16 Dimensions (L x W x H) 30 x 30 x 23 mm 30 x 30 x 23 mm 30 x 30 x 23 mm 30 x 30 x 23 mm Weight 20 1 g 20 1 g 20 1 g 20 1 g Mechanical Parameter AI VisionCube S AI VisionCube D AI VisionCube ST AI VisionCube DT Dimensions (L x W x H) 38 x 38 x 27.7 mm 38 x 38 x 26.7 mm 38 x 38 x 27.7 mm 38 x 38 x 26.7 mm Mounting hole 25.5 x 25.5 mm 25.5 x 25.5 mm 25.5 x 25.5 mm 25.5 x 25.5 mm Visible Light 1 Camera Parameter AI VisionCube S AI VisionCube D AI VisionCube ST AI VisionCube DT Sensor (inch) 1/2.8 1/2.6 1/2.8 1/2.6 Focal length (mm) 4 3.9 4 3.9 FOV 69 (H)x42 (V) 72 (H)x45 (V) 69 (H)x42 (V) 72 (H)x45 (V) Minimum illumination 7341 mV/luxs 9650 mV/luxs 7341 mV/luxs 9650 mV/luxs Video format 1920x1080@30hz 1920x1080@60hz 1920x1080@30hz 1920x1080@60hz Camera size (L x W x H) 19 x 19 x 30 mm 40.8 x 25 x 26 mm 19 x 19 x 30 mm 40.8 x 25 x 26 mm Visible Light 2 Camera Parameter AI VisionCube S AI VisionCube D AI VisionCube ST AI VisionCube DT Sensor (inch) / 1/2.6 / 1/2.6 Focal length (mm) / 12 / 12 FOV / 26 (H)x15 (V) / 26 (H)x15 (V) Minimum illumination / 9650 mV/luxs / 9650 mV/luxs Video format / 1920x1080@60hz / 1920x1080@60hz Camera size (L x W x H) / 40.8 x 25 x 26 mm / 40.8 x 25 x 26 mm Thermal Imaging (AI VisionCube ST / DT) Parameter AI VisionCube ST AI VisionCube DT Pixel size 12um 12um Focal length (mm) 9.1 9.1 FOV 20.3 (H)x15.2 (V) 20.3 (H)x15.2 (V) Spectral range 8~14um (LWIR) 8~14um (LWIR) Video format 384x288@25hz 384x288@25hz Camera dimensions (L x W x H) 26 x 26 x 42.63 mm 26 x 26 x 42.63 mm Interface USB USB Whats IncludedPackage list (AI VisionCube DT): AI VisionCube DT x 1 Camera Base x 1 SH1.0-12PIN 15cm Cable x 1 (for connection between the board and AI VisionCube Base) Cross Pan Head Screws M2*4 x 4 Hex Screws M2*6 x 4 Compatibility & Setup Notes Integration requirements (wiring, mounting, and software setup) depend on the drone platform and selected model. For custom projects or bulk purchases, contact customer service at support@rcdrone.top or https://rcdrone.top/. Applications FPV drone integration requiring onboard recognition, tracking, and OSD overlay Industry-grade 6-inch racing drones and ultra-compact cinewhoops (installation and tuning may be required) Visible + thermal workflows (ST/DT variants) for day/night operation in supported temperature ranges Details Droneer AI VisionCube brings onboard perception to FPV drones with recognition, tracking, and target-lock features plus OSD support. A modular, lightweight design integrates the camera system and AI image processing for fast installation on compact builds. Core functions include long-range recognition, multi-target detection, in-frame picture-in-picture, and adaptive tracking. Long-range recognition is designed for aerial scenarios where early identification and stable tracking matter. In-frame tracking keeps a wide landscape view while using an inset zoom window to maintain target awareness. YOLOv7-based detection supports multiple targets in one scene for dynamic environments. Adaptive tracking prioritizes the center-frame target for quick lock-on during high-speed movement. Dynamic memory lock-on helps re-capture a target after brief occlusion or when it leaves the frame. Thermal variants support visible and infrared sensor switching to extend perception across lighting conditions. The AI flight controller processes tracking data and helps coordinate autonomous tracking with manual control. Compare model options at a glance to match camera configuration and recognition range to your build. Detailed camera specs cover visible-light lens options and thermal imaging parameters where applicable. Each kit includes the VisionCube module, camera base, connection cable, and mounting hardware.