On November 26, 2022, the China Automation Congress 2022 (CAC) opened at the Xiamen International Conference‑Center Hotel in Fujian Province. CHINGMU Vision’s core R&D team originated from the Institute of Automation, Chinese Academy of Sciences. Zhang Haiwei, CEO of CHINGMU Vision, obtained his master’s degree from the State Key Laboratory of Pattern Recognition and Artificial Intelligence, Institute of Automation, CAS. His research covers SLAM localization, 3D reconstruction, panoramic stitching, motion capture and facial capture, with three authorized invention patents under his name. Backed by solid‑core technology from the Institute of Automation, CAS, CHINGMU has achieved breakthroughs in soft‑object tracking, large‑volume capture, multi‑person capture for over 40 performers, outdoor capture and marker‑less capture. The company attended the congress as a member of the Chinese Association of Automation.

CAC 2022 China Automation Congress

As a top‑tier industry‑wide academic event, CAC gathered leading experts from multiple automation‑related fields. For a long time, the optical‑motion‑capture market and product R&D have been monopolized by overseas brands. The high‑precision MC‑Series optical motion‑capture system exhibited by CHINGMU is fully self‑developed based on CAS‑originated technology, representing genuine “Made‑in‑China” innovation. Distinct from bad industry practices such as re‑branding and repackaging outdated foreign technology as new products, CHINGMU’s independently‑developed solution drew widespread attention from specialists. Face‑to‑face on‑site exchanges enabled experts to learn full‑cycle details of domestic optical‑motion‑capture workflows, covering both production and operation. The debut of the MC4000W camera became one of the hottest highlights of the exhibition.

MC4000W

As a high‑precision optical motion‑capture system, the MC4000W adapts to a broad spectrum of scenarios and environmental requirements. Its strengths of high accuracy, dead‑zone‑free tracking and ultra‑low‑latency satisfy diverse application‑demands.

Live‑Demonstration Site

The outstanding performance demonstrated by CHINGMU Vision’s motion‑capture system won high recognition from experts and student.

1. Soft‑Body Capture

Deformable soft bodies contain countless degrees of freedom, while rigid‑body objects maintain fixed geometric relationships between markers, whose six‑degree‑of‑freedom pose can be calculated with three or more markers. Soft‑body tracking requires a denser arrangement of markers to approximate deformation, with short inter‑marker distances; each marker must be computed independently, and its ID needs to remain consistent throughout capture.

One major advantage of active‑light tracking is permanent marker‑ID retention. Unique flashing sequences emitted by active markers deliver exclusive blob‑extraction results, ensuring stable marker‑ID recognition throughout recording. Even if markers are temporarily occluded and rediscovered, their original IDs can be restored. Permanent‑ID tracking enables clear differentiation between multiple identical‑looking objects. For example, during red‑blue confrontation training, every visually‑identical firearm can be assigned an independent ID for separate data analysis.

Passive‑marker systems commonly reset IDs after occlusion recovery, requiring tedious manual correction when large batches of tracking points are lost. The fixed‑ID feature of active‑light capture eliminates manual post‑processing, cuts labour‑and‑time costs and reduces human‑induced errors, guaranteeing high‑accuracy experimental‑data results.

Active‑light tracking also features strong anti‑interference capability and compact hardware size. Thanks to high‑luminance marker emission, signals maintain excellent penetration in bright outdoor environments or smoke‑filled fire‑drill training grounds, delivering powerful noise‑resistance and reliable detection.

2. Advanced‑Occlusion‑Resistance Algorithm

CHINGMU demonstrated domino‑effect capture on‑site. At the moment falling dominoes covered markers and rigid‑body targets overlapped, the system maintained stable, interference‑free high‑precision tracking even under heavy occlusion.

Position‑calibration workflows align rigid‑body orientation with the digital model, and self‑developed algorithms centre‑match 3D models to physical‑object coordinates. Even when markers are placed inconsistently (as seen across separate domino tiles), the physical‑world object remains perfectly registered to its 3D digital counterpart.

CHINGMU’s motion‑capture platform supports soft‑body, rigid‑body and full‑human capture. Its tracking zone extends over thousands of square meters, with a maximum tracking‑distance of 40 meters. Real‑time interactive capture can accommodate more than 40 participants, surpassing the current Guinness World Record for simultaneous human‑subject capture, and supports complex multi‑person interactions including hugs, running, jumping and overlapping poses.

3. 89°×89° Wide Field‑of‑View for Dead‑Zone‑Free Capture

During the CAC demonstration within a 3 m × 3 m testing area, the MC4000W (89°×89° field‑of‑view) tracked rigid‑body targets moving to every corner, even under partial occlusion. Remarkably, objects temporarily beyond the camera’s nominal detection boundary could still be captured reliably, impressing on‑site experts and students. This demonstrated the powerful dead‑zone‑free tracking capability of the MC4000W.

SLAM‑Algorithm‑Verification for Unmanned‑Vehicles, Wenzhou University

The versatile MC‑Series suits a wide range of research‑oriented applications, including UAV trajectory tracking, drone‑vehicle cooperative experiments, algorithm validation, robotic‑arm high‑precision positioning and robot pose tracking. CHINGMU Vision has established research partnerships with Shanghai University, Wenzhou University, Kunming University of Science and Technology and other academic institutions. Further information is available via the official website and WeChat public account of CHINGMU Vision.