RVC M2600C Built for Precision Segmentation and Picking

1440×1080 binocular color imaging captures color, texture, and 3D spatial information together. Optimized for highly reflective, multi-reflective, and black light-absorbing surfaces, it supports instance segmentation, precision picking, agricultural harvesting, and object detection.

RVC M2600C RVC M2600C mounted at a robot end-effector

01 / Key Challenges

Difficult to Segment Similar Parts

Stacked, closely spaced, or partially occluded parts with similar geometries can be difficult to distinguish reliably using contour information alone.

02 / Key Challenges

Specular Reflections on Metal Surfaces

Beveled edges, sharp corners, and complex metal surfaces cause strong specular reflections and multiple reflection paths, resulting in incomplete point clouds and distorted object boundaries.

Robot handling a black glossy automotive panel

03 / Key Challenges

Black and Light-Absorbing Surfaces

Black-painted surfaces and dark materials produce weak sensor returns, sometimes accompanied by localized specular reflections, resulting in incomplete depth data and reduced accuracy in segmentation and grasp pose estimation.

agricultural robot harvesting fruit among foliage

04 / Key Challenges

Harsh Lighting and Foliage Occlusion

Fruits are partially or fully occluded by leaves and branches, while constantly changing natural light makes it challenging to maintain stable 2D recognition and accurate depth localization simultaneously.

Product Features

Robot capturing a color image and generating a 3D point cloud — Typical Point Cloud Examples/ Part Instance Segmentation

Acquisition scene for a highly reflective metal workpiece — Typical Point Cloud Examples/ Highly reflective Metal Parts

Point-cloud result for a highly reflective metal workpiece — Typical Point Cloud Examples/ Highly reflective Metal Parts

Black glossy painted workpiece — Typical Point Cloud Examples/ Black and Light-Absorbing Surfaces

Point-cloud result for a black glossy painted workpiece — Typical Point Cloud Examples/ Black and Light-Absorbing Surfaces

Outdoor tree color image and 3D point-cloud result — Typical Point Cloud Examples/ Outdoor Natural Lighting

Built for Precision Segmentation and Picking High-Resolution Color Imaging

With 1440 × 1080 binocular color imaging, the system captures color, texture, and 3D spatial information simultaneously, providing richer visual features for object detection and instance segmentation.

Highly Reflective Workpieces Fine Point Clouds

The multi-reflection suppression algorithm minimizes interference from complex reflective surfaces, improving point-cloud completeness and accuracy around grooves, sharp edges, and complex geometries.

Dark and Light-Absorbing Surfaces Robust 3D Imaging

Optimized for black, light-absorbing surfaces, localized reflections, and multiple-reflection conditions, the 3D imaging system delivers more complete point clouds on dark-colored parts and complex material surfaces.

Line-Scan Mode with High Ambient-Light Resistance >60 0,000 Lux

Supporting natural daylight, semi-outdoor environments, and high-intensity ambient lighting, the system delivers stable and reliable 3D imaging for agricultural harvesting and open workstation applications.

Typical Applications

Weld Seam Detection

Part Segmentation

Random Bin Picking

Color-Based Classification

Agricultural Harvesting

Object Detection

Defect Detection

3D Localization

Product Specifications

RVC M2600C working distance from 400 to 1000 mm and field-of-view diagram

RVC-M2600C Parameter Table
Category	Parameter Item	RVC-M2600C
Basic Data	Camera Dimensions	170 × 73.5 × 73 mm
	Camera Weight	1.2 kg
	Camera Resolution	1440 × 1080 (1.6 MP), binocular, color
	Light Source	Laser
	Data Interface	Gigabit Ethernet
	Operating Voltage / Current	DC 24V / 2A
	SDK Supported Systems	Linux / Windows
	SDK Supported Languages	C / C++ / C# / Python
Performance	Working Distance	400–1000 mm
	Near FOV	365 × 231 mm @ 400 mm
	FOV at Optimal Working Distance	600 × 450 mm @ 700 mm
	Far FOV	924 × 640 mm @ 1000 mm
	Z-Axis Single-Point Repeatability	0.04 mm @ 1000 mm
	Z Area Repeatability	0.02 mm @ 1000 mm
	XYZ Spatial Measurement Error	0.27–0.66 mm；full FOV <0.2%；FOV center <0.1%
	Capture Time	1.1–2.3 s / frame
	Measurement Notes	Performance values tested at a normal room temperature of 26℃
Stability	Housing Protection	IP65 Ingress Protection
	Ambient-Light Resistance	Line-scan mode >600,000 Lux
	Operating Temperature	-20–50℃