matlab-read-write-point-cloud-file

Use lasFileReader to create a reader object, then call readPointCloud on it. The two-output form [ptCloud, ptAttributes] = readPointCloud(reader) also returns classification, GPS timestamps, and other lidar attributes. Calling pcread on a LAS file throws an error.

Use velodyneFileReader with the file path and a mandatory device model string such as "HDL32E" or "VLP16". Then loop with hasFrame and readFrame to iterate through frames, since there is no function to load all frames at once.

No, pcwrite only supports PLY and PCD formats. For LAS or LAZ output, create a lasFileWriter object and call writePointCloud. Pass the attributes object as a third argument to preserve classification and GPS timestamps.

PLY format only supports "ascii" and "binary" encoding, so Encoding="compressed" throws an error. The compressed encoding is exclusive to PCD files, where it is the default in R2026a and later.

Classification, GPSTimeStamp, LaserReturn, and ScanAngle live on the lidarPointAttributes object returned as the second output of readPointCloud, not on pointCloud. The pointCloud object holds Location, Color, Normal, and Intensity.

Do not use file readers for streaming live sensor data in real time; use velodynelidar, ousterlidar, or sicklidar instead. File readers are also not for post-read processing like downsampling or denoising, which use pcdownsample and pcdenoise.