Grape detection, segmentation and tracking using deep neural networks and three-dimensional association.
Thiago T. SantosLeonardo L. de SouzaAndreza A. dos SantosSandra AvilaPublished in: CoRR (2019)
Keyphrases
- detection responses
- neural network
- three dimensional
- occlusion reasoning
- complex background
- automatic initialization
- motion segmentation
- appearance model
- data association
- accurate localization
- cluttered background
- reliable detection
- geometric active contours
- bounding box
- false detections
- foreground background separation
- human detection
- pattern recognition
- curvilinear structures
- text detection
- image segmentation
- multiscale
- joint detection
- segmentation method
- detection method
- segmentation algorithm
- partial occlusion
- multi target tracking
- background modeling
- artificial neural networks
- multiple objects
- image sequences
- detection algorithm
- range images
- video scene
- face detection and tracking
- level set
- low signal to noise ratio
- medical images
- detecting and tracking multiple
- surface representation
- computer vision
- multiple cues
- motion model
- object tracking
- mean shift
- active contours
- orientation estimation
- particle filtering
- detecting objects
- phase contrast
- video surveillance
- shape prior
- visual tracking
- medical imaging
- ellipse fitting
- x ray
- particle filter
- object detection