Grape detection, segmentation, and tracking using deep neural networks and three-dimensional association.
Thiago T. SantosLeonardo L. de SouzaAndreza A. dos SantosSandra AvilaPublished in: Comput. Electron. Agric. (2020)
Keyphrases
- detection responses
- neural network
- three dimensional
- occlusion reasoning
- automatic initialization
- reliable detection
- motion segmentation
- accurate localization
- foreground background separation
- cluttered background
- complex background
- geometric active contours
- image segmentation
- curvilinear structures
- multi target tracking
- multiscale
- low signal to noise ratio
- data association
- detection algorithm
- appearance model
- level set
- object detection
- false detections
- detection method
- video scene
- text detection
- face detection and tracking
- medical images
- pattern recognition
- bounding box
- human detection
- segmentation method
- mean shift
- joint detection
- object segmentation
- particle filtering
- partial occlusion
- detecting and tracking multiple
- rib cage
- surface representation
- visual tracking
- energy function
- artificial neural networks
- image sequences
- detecting objects
- multiple cues
- multiple objects
- detection rate
- background subtraction
- ellipse fitting
- motion model
- range images
- d objects