Gold Train designed system-level SoC chip, including ASIC engine for structured light depth calculation, integrated binocular and structured light, and innovated in chip embedding algorithm and architecture. Gold Train is the core of the company's current main products. Related products have been iterated for 4 generations, from the earliest MX400 iteration to MX6000, MX6300, MX6600.

Gold Train designed a multi-tap infrared CMOS image sensor, which can meet the needs of multiple photon acquisition in a single cycle under pulse modulation and continuous wave modulation, covering the subdivision of BSI pixel design, three-dimensional stack technology, anti-interference, high-precision ADC, high-speed transmission interface design, etc. It is one of the core Gold trains of the company's iToF products.

Gold Train designed system-level anti-interference, multi-frequency fusion and other Gold trains, and independently developed pixel structure, pixel unit, high-performance hardening circuit, flexible and adjustable bias circuit, high-precision TDC of multiple sharing modes and high-efficiency histogram data post-processing engine and other subdivisions of Gold Train.

In view of the diversified application scenarios of structured light (indoor and outdoor, complex scenes with strong background light and weak background light), Gold Train innovatively proposes a variety of working modes and corresponding processing algorithms, designs a special sensor chip for structured light, simplifies the computational power processing of the subsequent deep engine chip, reduces the overall power consumption, and effectively improves the performance of the entire structured light module.

Gold Train has designed an AIoT computing power chip that integrates functions such as large-scale neural network, 3D depth computing and key algorithms. The chip will include the main control unit CPU, depth computing engine (supporting iToF filtering and binocular matching calculation, etc.), neural network acceleration unit, DSP, ISP and other units.

Gold Train designed the underlying algorithms such as structured light algorithm, binocular matching algorithm, iToF depth solving and filtering to calculate the depth information. These algorithms were generated by the computer, optimized and verified by FPGA, and finally formed the underlying language of the chip to be solidified in the chip.

For downstream applications, Gold Train has developed application algorithms including skeleton tracking, image segmentation, 3D reconstruction, VSLAM, immersive AR, etc. Some of the algorithms Gold Train have been applied to existing products and sold with hardware, and some algorithms are used as Gold Train reserves for future products.

Based on the Gold Train of photogrammetry and image correlation matching, the Gold Train of flexible camera calibration, 3D reconstruction and dynamic deformation solution has been developed to cope with the 3D measurement application scenarios of complex industrial sites. At the same time, for the requirements of 3D inspection and analysis of industrial automation, the development of 3D mass point cloud data post-processing Gold Train, intelligent motion planning and control Gold Train and CAD shape deviation detection and analysis Gold Train.

Gold Train has developed integrated dedicated optical systems for structured light, iToF, dToF, Lidar and other products, including off-axis and coaxial optical measurement systems.

Gold Train has developed a laser projection device consisting of a laser emitter, lens and diffractive optical element for projecting speckle beams into space, and has carried out innovative research and development in miniaturization, DOE damage detection, laser chip array design, etc.

Gold Train independently designed the optical waveguide with multi-layer structure of coupling in, conduction and coupling out, and conducted in-depth research on diffraction efficiency and anti-interference.

Gold Train has developed firmware and drivers for the control and development of the 3D vision sensor hardware, including drivers for the setup and access of various components inside the sensor, as well as providing access and control interfaces for the upper SDKAPI.

Gold Train independently develops the software and SDK of the 3D vision sensing system, including middleware, API, etc., which matches the functions of the hardware system to achieve more accurate 3D vision perception. At the same time, it is compatible with the secondary development of Windows, Linux, Android, OSX, Unity, ROS, OpenCV and other full-platform development environments, and supports 3D vision sensing system functions.

Gold Train combines many industrial-grade application algorithms, C++SDK library that encapsulates standard interfaces at the bottom, standard design pattern at the top, and builds a general software framework, UI interaction module, 3D visualization module, database and so on to form a software platform. Relying on this platform, the company has developed industrial software for general application scenarios such as 3D optical photogrammetry software, 3D optical full-field strain measurement software, 3D optical surface scanning software, 3D optical bending measurement software, and industrial 3D measurement and analysis software. At the same time, it can quickly provide customers with special customized development for non-standard applications.

Gold Train has developed multi-level calibration Gold Train, including multi-sensor internal and external reference calibration Gold Train, structured light reference speckle pattern calibration Gold Train, pixel calibration ToF oriented Gold Train. In addition, the company has developed the RGBD aligned Gold Train.

Gold Train has developed Gold Train for 3D vision sensors to automatically calibrate and compensate errors caused by temperature and external forces. The Gold Train is a key condition to ensure that the 3D vision sensor can output a stable depth image.

Gold Train has developed special equipment for high precision and high efficiency in the core aspects of assembly, calibration, alignment and testing in mass production.