The Percipio depth camera, using its own proprietary active stereo technology, consist of the stereo infrared image sensor, color image sensor and laser projector.
The camera receives software commands from the host computer through the USB or Ethernet interface to capture images. Cameras with a hardware trigger interface support external hardware signal triggering, and capture images based on the frequency of the trigger signal (≤ frame rate of continuous mode).
After the laser projector casts structured light onto the surface of the object being measured, the infrared image sensor receives reflected infrared data from the two channels and the color image sensor captures RGB data. With an embedded vision processor as the computing platform, the camera rapidly calculates the depth data of each pixel within the field of view based on the principle of triangulation, and then uploads the depth data, infrared data, and RGB data to the host computer via the interface.
The Percipio depth camera, using Time of Flight (ToF) technology, consist of the color image sensor, ToF sensor and laser projector.
The camera receives software commands from the host computer through the Ethernet interface to capture images. Cameras with a hardware trigger interface support external hardware signal triggering, and capture images based on the frequency of the trigger signal (≤ frame rate of continuous mode).
After the laser projector casts modulated light onto the surface of the object being measured，the ToF sensor receives the modulated light reflected from the surface of the object. The color image sensor captures RGB data. With an embedded vision processor as the computing platform, the camera rapidly calculates the depth data of each pixel within the field of view. Finally, the depth data, infrared data, and RGB data are uploaded to the host computer via the camera interface.
The depth map is a 16-bit single-channel matrix, composed of depth data of all pixels in the depth camera’s field of view. To visualize the distance values, all SDK sample programs map the output depth map to the RGB color space, resulting in an 8-bit three-channel bitmap.
The individual pixel value in the depth map is the vertical distance in millimeters from a point on an object in space to the plane perpendicular to the optical axis of the left infrared lens and passing through the optical center of the lens (depth camera optical zero point). Points without depth data have a value of 0. The depth data only includes intrinsic parameters for point cloud conversion, not extrinsic parameters. The depth data output from an active stereo camera is undistorted, while the depth data output from a ToF camera is distorted.
Point clouds are a data matrix composed of the point cloud information of all points within the field of view of a depth camera. The point cloud information of each point consists of three-dimensional coordinates (x, y, z). Points without three-dimensional spatial information are represented as (x, y, 0).
Infrared image (IR image) is an image output by an infrared image sensor. Some models of Percipio depth cameras output infrared images that have been processed by the system. If you need to view original infrared images, you need to disable the depth map output. Infrared images are divided into left infrared images and right infrared images, both of which contain intrinsic and distortion parameters. However, because the left infrared images and the depth maps are in the same spatial coordinate system, the left infrared images have no extrinsic parameters. Active stereo cameras output both left and right infrared images, while ToF cameras only output the left infrared images.
Color image (RGB image) is an image output by a color image sensor. The color image sensor component provides intrinsic parameters, extrinsic parameters, and distortion parameters. Different models of Percipio depth cameras will output color images in various formats.
The color image sensor with hardware ISP module outputs YUYV422/JPEG images, which can be processed by OpenCV to display color images in the standard color space.
The color image without a hardware ISP module outputs RAW Bayer images, which may have color cast. The software ISP of the SDK processes the raw bayer image (e.g., white balance) and displays it as a color image in the standard color space. A color image sensor without a hardware ISP module can ensure that the output image data is synchronized with the infrared image data.
