realsenseai/librealsense

Features

• Depth Estimation - Computes depth maps and spatial data using an internal vision processor to reduce host CPU load.
• Depth Camera SDKs - Provides a comprehensive SDK for interfacing with depth cameras to capture color, depth, and infrared data.
• High Precision Depth Sensing - Measures object distance with sub-millimeter accuracy at short ranges using a stereo vision system.
• Camera Calibration - Injects calibration data to ensure millimeter-level accuracy of captured depth measurements.
• Stereo Calibration - Automatically calibrates stereo cameras in seconds without requiring the use of specialized targets.
• Point Cloud - Transforms depth and infrared data into 3D point sets to represent physical space.
• Point Cloud Generators - Transforms depth and infrared frames into three-dimensional spatial representations.
• Camera Data Ingestion - Provides language bindings and framework wrappers to ingest depth and color streams for easier processing.
• Depth-Color Alignment - Synchronizes RGB data with depth maps using an image signal processor to ensure aligned visual information.
• Real-time Depth Streaming - Captures real-time depth and color information at specified resolutions and frame rates.
• Stereo Vision Reconstruction - Computes depth maps and estimates movement using an on-chip vision processor and digital signal processor.
• Peripheral Device Managers - Discovers connected devices, retrieves hardware specifications, and performs remote resets over a network.
• Hardware Device Discovery - Fetches hardware details such as serial numbers and firmware versions via dedicated service calls.
• Sensor Drivers - Provides a hardware abstraction layer for device discovery and raw data ingestion from depth sensors.
• Cross-Language Bindings Layers - Exposes low-level C++ hardware interfaces to higher-level languages via standardized wrapper libraries.
• Depth Frame Filtering - Aligns depth data to color frames and applies filters to refine the resulting images.
• Pose Estimation - Extracts spatial orientation and positioning information from pose frames.
• Robot Operating System Wrappers - Exposes camera data and IMU measurements as topics and services for robot operating systems.
• High Resolution Sensor Streaming - Captures depth and color imagery using global shutter sensors to minimize motion blur.
• Camera Topic Exposure - Exposes camera data as topics and services for seamless integration within a ROS 2 ecosystem.
• Data Streaming Interfaces - Streams data to nodes via a compatible interface to avoid installing additional software in robot operating system ecosystems.
• Concurrent Device Operation - Connects and runs several depth sensors simultaneously without causing signal interference.
• Hardware Stream Synchronization - Aligns data from multiple sensors using physical sync cables and hardware triggers for precise timing.
• Image Buffer Sharing - Transforms image data into arrays using a shared buffer protocol to avoid expensive memory copying.
• Sensor Data Publishing - Streams accelerometer and gyroscope samples including metadata and synchronization information.
• Virtual Hardware Simulators - Creates software-defined camera devices to emulate physical sensor behavior for testing and development.
• Depth Accuracy Metrics - Evaluates the precision and fill rate of distance data to verify performance in specific environments.
• Geometric Distortion Correction - Applies geometric distortion correction and temporal noise reduction to improve imagery quality.
• Real-time Point Cloud Streaming - Sends frame timestamps, metadata, and 3D point cloud data to clients using a persistent socket connection.
• RGB and Depth Synchronization - Synchronizes color and depth images into a single message for simplified spatial analysis.
• Distributed Clock Synchronizers - Aligns data streams from multiple cameras using hardware links or external sync cables for precise timing.
• Inertial Measurement Unit Interfaces - Accesses an onboard IMU to track camera movement and refine environmental awareness.
• Inertial Environmental Tracking - Uses an inertial measurement unit to refine depth awareness and environmental tracking during movement.
• Inertial Frame Alignment - Aligns spatial orientation with depth frames using time-stamped inertial measurement unit data.
• Spatial Orientation Tracking - Measures spatial movement and orientation using a built-in inertial measurement unit.
• Sensor Configuration - Provides tools to update sensor options like exposure and gain and manage stream profiles.
• Data Compression - Reduces bandwidth usage by applying compression plugins to color, depth, and infrared data streams.
• Vision Stream Compression - Applies modular compression algorithms to color and depth data to reduce network bandwidth usage.
• Real-time Volumetric Captures - Coordinates data from several sensors simultaneously for real-time volumetric scene capture and 3D scanning.
• Buffer-Protocol Array Integration - Transforms depth and color image data into NumPy arrays using the buffer protocol for zero-copy transfers.
• Zero-Copy Mechanisms - Reduces latency and CPU overhead for high-resolution frames by using zero-copy communication between internal processes.

Historique des stars