Classify Ground by Deep Learning

Function Overview

Uses a deep learning model to classify ground points within point cloud data. Points from the initial class that meet the criteria will be classified into the target class. Points not classified as the target class will retain their original class value.

Usage

Click Classify > Classify Ground by Deep Learning。

Parameter Settings

• From Class：The class of points to be classified.
• ​​Ground Points Class: Ground Points.
• To Class：The target class for the classification results. Default is "2-Ground".
• Default：Click this button to restore all parameters to their default values.
• Use GPU first：Whether to use GPU acceleration for classification. This function supports two operation modes: GPU and CPU. If the computer's graphics card meets the requirements (NVIDIA GPU with compute capability 3.5 or above, and at least 4GB of available VRAM at runtime), this option will be selected by default. Users can choose whether to use the GPU based on their needs. GPU operation is approximately 4 times faster than CPU. If the conditions are not met, the GPU cannot be used, and the function will default to running on the CPU.

  Accuracy Metrics

The model achieves an accuracy of 94.66%.

Note:

This function's output will overwrite the original data files.