Image stacking versus per-frame results combination for video text recognition.
Boris SavelyevVladimir L. ArlazarovPublished in: ICMV (2020)
Keyphrases
- input image
- image data
- text recognition
- key frames
- video frames
- image content
- template matching
- multiscale
- image classification
- image features
- video images
- image retrieval
- image segmentation
- image collections
- image representation
- region of interest
- video sequences
- multimedia
- segmentation method
- image regions
- object detection
- hough transform
- segmentation algorithm
- language model
- lighting conditions
- object recognition
- bounding box
- complex scenes
- viterbi algorithm