![]() For details, please see drivers/zed_capture/README.md. We provide a Python tool for generate LDC rectification tables for ZED stereo camera. In order to use LDC, the rectification tables should be provided in the format that LDC supports. LDC (Lense Distortion Correction) Īs shown in Figure 1, we use the LDC HWA to rectify the left and right images. Each point in the point cloud represents a 3D position (X, Y, Z) and includes color information (R, G, B).Ħ.4.1. The overall application data flow is shown in Figure 1.Īdditionally, when configured, the output disparity map and the rectified right image can be mapped to generate a 3D point cloud through a triangulation process. Up to 3 layers are supported and these are configurable. In the multi-layer SDE refinement mode, it combines the disparity maps produced by the SDE at different layers with post processing. In the single-layer SDE mode, it outputs the raw disparity map from the SDE without any post-processing. The application supports two disparity estimation modes: the “single-layer SDE mode” and the “multi-layer SDE refinement mode”. The SDE produces a disparity map from the rectified stereo images. It’s important to provide the rectification tables in a format that the LDC can recognize. The LDC converts the input stereo images to YUV420 (NV12) format and rectifies the images using two rectification tables, one for the left camera and one for the right camera. The input image format for this application is YUV422:UYVY. The application not only outputs the raw disparity map but also generates a point cloud with 3D position (X, Y, Z) and color information (R, G, B). This demonstrates the stereo application that utilizes the LDC (Lens Distortion Correction) and DMPAC SDE (Stereo Depth Engine) hardware accelerators (HWAs). Stereo demo: disparity and confidence maps
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |