Hi computational photography and vision researchers,
I've been interested in this category of problems in a specific setting and I'd like to have a sense of the broader fields with relevance. Is there any work on image restoration or image enhancement that the image is not only optimized for better visual quality, but also towards better CV tasks such as object recognition?
Any pointer would be appreciated!
Pattern recognition and computer vision usually deals with a pre-processing step where the image is enhanced in quality for better feature extraction. Steps involved are tonal correction, noise reduction, morphological operations and segmentation.
Hi Zihao W. Wang
I once read a paper which aims to super resolve a video for the better performance of action recognition: "Zhang, Haochen, Dong Liu, and Zhiwei Xiong. Two-stream action recognition-oriented video super-resolution. ICCV 2019". Hope it helps.
Best,
Yongcheng
Image classification and computer vision generally require a pre-processing phase where the image is improved inconsistency for the excellent combination of the functionality. Tonal modification, noise reduction, morphological processes, and standardization are all concerned.
Yongcheng Jing , thank you! This is exactly the type of work I'm thinking of. I appreciate it!
- Badges
- Science topic
- Similar topics
- Computer Science
- Images