Main Content

Segmentation and Detection Using Deep Learning

Semantic segmentation and object detection in point cloud data using deep learning algorithms

Semantic segmentation associates each point in a 3-D point cloud with a class label, such as car, truck, ground, or vegetation. Lidar Toolbox™ provides deep learning algorithms to perform semantic segmentation on point cloud data. Use PointSeg and SqueezeSegV2 convolutional neural networks (CNN) to develop semantic segmentation models.

Lidar Toolbox also provides the object detection CNN PointPillars for developing custom object detection models.

Functions

expand all

combineCombine data from multiple datastores
countEachLabelCount occurrence of pixel or box labels
imageDatastoreDatastore for image data
pixelLabelDatastoreDatastore for pixel label data
boxLabelDatastoreDatastore for bounding box label data
fileDatastoreDatastore with custom file reader
transformTransform datastore
randomAffine3dCreate randomized 3-D affine transformation
bboxwarpApply geometric transformation to bounding boxes
squeezesegv2LayersCreate SqueezeSegV2 segmentation network for organized lidar point cloud
semanticsegSemantic image segmentation using deep learning
labeloverlayOverlay label matrix regions on 2-D image
showShapeDisplay shapes on image, video, or point cloud
pcshowPlot 3-D point cloud
evaluateSemanticSegmentationEvaluate semantic segmentation data set against ground truth
evaluateDetectionAOSEvaluate average orientation similarity metric for object detection
bboxOverlapRatioCompute bounding box overlap ratio

Topics

Getting Started with Point Clouds Using Deep Learning

Understand how to use point clouds for deep learning.

Featured Examples