Intel RealSense D435 Depth Camera for Fast Wide‑FOV 3D Scanning and Tracking
The Intel RealSense D435 is a wide‑field‑of‑view depth camera built around the D430 stereo depth module and Vision Processor D4. With an industry‑leading 87° × 58° depth FOV and a 50mm stereo baseline, it captures dense 3D point clouds at up to 1280 × 720 resolution and 90 fps — making it the go‑to choice for fast‑moving robotics, autonomous navigation, and real‑time 3D scanning tasks. The D435 uses global shutter depth sensors for sharp depth capture even under motion, and pairs it with a 2 MP rolling‑shutter RGB camera (1920 × 1080 @ 30 fps) for colour overlay. It draws power over a single USB‑C 3.1 Gen1 cable (~2.5 W typical) and runs on the open‑source librealsense SDK 2.0 with full ROS / ROS 2 support. Features Wide FOV Depth Camera — 87° × 58° depth field of view and 1280 × 720 at up to 90 fps give you the widest real‑time depth coverage in the RealSense lineup, ideal for close‑range room‑scale and desktop 3D capture. Active Stereo Vision — D430 stereo module with global shutter IR imagers projects a coded IR dot pattern to compute depth robustly across textured, semi‑textureless, and indoor surfaces. 50 mm Stereo Baseline — Compact baseline keeps the camera narrow (90 mm) while still delivering reliable depth out to ~3 m indoor. Sub‑centimetre Depth Accuracy — < 2% error at 2 m; Min‑Z depth starts at ~28 cm at full resolution, enabling fine‑grained close‑range 3D scanning. Global Shutter Depth Sensors — Eliminates motion blur in moving‑scene and robot‑arm capture scenarios, critical for dynamic environments. 2 MP RGB Camera — 1920 × 1080 rolling‑shutter colour camera at 30 fps with 69° × 42° FOV for texture‑mapping and colour‑point‑cloud generation. No IMU — No gyroscope/accelerometer on board; choose the D435i if your project requires 6‑axis IMU data alongside depth. Vision Processor D4 — Dedicated depth‑processing ASIC handles all stereo matching on‑board, offloading host CPU and reducing USB bandwidth requirements. USB‑C Single‑Cable Power & Data — Bus‑powered via USB‑C 3.1 Gen1; ~2.5 W typical power draw, no external adapter needed. Dual Mounting Threads — 1/4‑20 UNC tripod thread plus 2× M3 screw holes for flexible robot‑frame, tripod, or arm attachment. On‑Device Calibration — < 15‑second on‑board self‑calibration; no calibration target required for most use cases. Open SDK & ROS Support — Cross‑platform librealsense 2.0 (Python, C++, C#, Java, MATLAB) and realsense‑ros packages for rapid prototyping and production deployment. Specifications Parameter Specification Depth Technology Active Stereo Vision (D430 Module) Image Sensor Type Global Shutter (depth); Rolling Shutter (RGB) Ideal Depth Range 0.3 m – 3 m (indoor); up to ~10 m in low‑ambient light Min‑Z (Min Depth Distance) ~28 cm at full 1280 × 720 resolution Depth Accuracy < 2% at 2 m distance Depth FOV 87° × 58° (±3° / ±2°) Depth Resolution / Max FPS 1280 × 720 @ 90 fps; 848 × 480 @ 120 fps; 640 × 480 @ 90 fps Stereo Baseline 50 mm RGB Sensor Resolution 1920 × 1080 (2 MP) RGB Frame Rate Up to 30 fps RGB FOV 69° × 42° RGB Sensor Technology Rolling Shutter Vision Processor Intel RealSense Vision Processor D4 IMU None (choose D435i for 6‑axis IMU) Power Supply USB‑C 3.1 Gen1; ~2.5 W typical (< 4 W peak) Connectivity USB‑C 3.1 Gen1 (USB‑C to USB‑A / USB‑C cable) Dimensions 90 mm × 25 mm × 25 mm Weight ~72 g (body only, cables not included) Operating Temperature 0°C – 35°C Mounting 1× 1/4‑20 UNC tripod thread + 2× M3 screw holes Reference Documents Intel RealSense D435 — Official Specifications librealsense SDK 2.0 — GitHub Repository D400 Series Datasheet — Intel RealSense Developer Site Application Robotics & Autonomous Navigation — Wide 87° FOV gives robots a broader situational awareness window for obstacle avoidance, path planning, and SLAM in indoor environments. 3D Scanning & Reconstruction — 1280 × 720 @ 90 fps depth + 2 MP RGB produce dense, colour‑mapped point clouds for desktop and room‑scale 3D capture on PC or embedded SBC. Augmented & Virtual Reality (AR / VR) — Fast global‑shutter depth tracking at up to 90 fps enables responsive hand‑tracking and environment‑mapping for immersive experiences. Gesture & Object Recognition — Coded‑dot IR pattern penetrates semi‑textureless surfaces, making the D435 reliable for interactive kiosks, signage, and smart‑shelf applications. Industrial Inspection & Quality Control — Sub‑centimetre depth accuracy and USB‑C simplicity suit factory‑floor 3D scanning and part‑verification setups. Drone‑Based Mapping — Lightweight 72 g body with tripod mount enables compact UAV payloads for low‑altitude mapping and terrain modelling missions. Education & Research Prototyping — Librealsense 2.0 with Python / ROS 2 bindings and < 15‑second on‑board calibration makes it ideal for academic labs and rapid prototyping courses. Logistics & Warehouse Automation — Wide FOV covers pallet‑width aisles efficiently; 90 fps depth supports moving conveyor‑belt scanning workflows. Packing List Description Quantity Intel RealSense D435 Depth Camera × 1 Stand × 1 USB-C Cable × 1 FAQ 1. What is the difference between the D435 and the D435i? The D435i adds an integrated 6‑axis IMU (Bosch BMI055) to the D435 hardware. If your application needs gyroscope and accelerometer data (e.g., sensor‑fusion for drone stabilisation, visual‑inertial odometry, or tilt‑compensation), choose the D435i. If you only need depth and RGB, the D435 offers the same optics and processing at a slightly lower price. 2. What is the effective depth range of the D435? The D435 performs best in the 0.3 m – 3 m indoor range. In low‑ambient‑light conditions it can reach ~10 m. Outdoor range is reduced due to sunlight interference with the IR dot projector. The Min‑Z (closest usable depth) is ~28 cm at 1280 × 720 resolution. 3. Does the D435 work outdoors? The D435 can be used outdoors in shaded or low‑ambient‑light conditions, but direct sunlight saturates the IR projector and degrades depth quality significantly. For permanently outdoor installations consider the D455 series with its longer 95 mm baseline and stronger IR illuminator, or use sun‑shading enclosures. The D435 is optimised for indoor and controlled‑environment use. 4. Why does the D435 use global shutter sensors? Global shutter captures all pixels simultaneously, eliminating the row‑by‑row exposure lag inherent to rolling‑shutter sensors. This makes the D435 ideal for moving‑scene capture — robot arms, drones, fast gestures — where rolling‑shutter would produce smeared or distorted depth data. 5. Can I use the D435 with ROS or ROS 2? Yes. The official realsense‑ros and realsense2‑ros packages expose depth, colour, and IMU (D435i) topics as standard sensor messages. Ubuntu 18.04 / 20.04 / 22.04 are officially supported. Pre‑built Docker images and snaps are also available for quick setup. 6. Can the D435 be powered over USB only? Yes. The D435 is bus‑powered via the USB‑C 3.1 Gen1 connection. Typical power draw is ~2.5 W (peak ~4 W). A USB‑C port on a desktop PC or a powered USB hub is sufficient. No external power adapter is required in most setups. 7. How does the D435 compare to the D455 series? The D455 series (D455, D455F, D456) features the newer D450 stereo module with a 95 mm baseline, deeper range (up to ~20 m), and optional features (IR Pass Filter on D455F, IP65 on D456). The D435 uses the compact D430 module with a 50 mm baseline, is narrower (90 mm vs 124 mm), lighter (~72 g vs ~133 g), and is better suited for desktop, close‑range, and budget‑constrained projects. 8. Does the D435 support multi‑camera synchronisation? The D435 supports hardware sync via the external sync cable header on the board. Multiple D435 units can be frame‑synchronised for 360° coverage, volumetric capture, or stereo‑pair setups. Software‑sync (timestamp matching) is also possible through the librealsense API. For advanced multi‑camera rigs, Intel's official Multi‑Camera Activation Guide provides wiring and timing details. 9. What programming languages and frameworks support the D435? The librealsense 2.0 SDK provides official bindings for Python, C++, C#, Java, and MATLAB. Community bindings also exist for Node.js, Unity (RealSense Unity package), Unreal Engine, and ROS/ROS 2. The Intel RealSense Viewer app gives you a zero‑code GUI for stream preview and recording. 10. Can the D435 be used for long‑range outdoor mapping? For long‑range (> 5 m) or permanent outdoor deployments, the D435's 50 mm baseline and IR projector power are limiting factors. Consider the D455 (longer 95 mm baseline, up to ~20 m range) or a dedicated LiDAR sensor like the Intel RealSense L515 for long‑range, sunlight‑immune scanning. The D435 excels at close‑ to mid‑range (0.3 – 3 m) indoor applications where its wide FOV and high frame rate are most advantageous.
