Training of a Convolutional Neural Network and Its Object Recognition Ability Depending on Illumination and Contrast of Images.
Dmytro KyrychukAndriy SeginPublished in: ACIT (2021)
Keyphrases
- object recognition
- image features
- illumination conditions
- image database
- image understanding
- lighting conditions
- imaging conditions
- test images
- convolutional neural network
- image matching
- three dimensional
- input image
- image data
- image classification
- enhanced image
- computer vision
- visual object recognition
- image analysis
- pixel intensities
- image retrieval
- object categories
- feature matching
- d objects
- image registration
- multiscale
- contrast enhancement
- illumination invariant
- varying illumination
- natural images
- edge detection
- illumination correction
- color constancy
- outdoor scenes
- machine learning
- feature selection
- transformed images