Abstract
Radon transform-based motion blurred silkworm pupa image restoration
Highlights
Sericulture industry has big importance in the growth of economic and people’s income
The kernel estimation accuracy of simulated data experiments was evaluated by kernel similarity and the quality of restored images were evaluated by peak signal-to-noise ratio (PSNR), Structural similarity (SSIM)
Blurred image applied with motion blur kernel parameter
Summary
Sericulture industry has big importance in the growth of economic and people’s income. The silkworm pupae image restoration was a blind deconvolution problem that the kernel needed to be accurately estimated first. Tao D, et al Radon transform-based motion blurred silkworm pupa image restoration. By predicting the explicit edges from the input blurred image, accurate kernel was estimated without using priors. Cho et al.[20] acquired sharp edges by adopting filtering techniques These methods imposed computation efficient Gaussian priors, resulting in the un-accurate kernel. B. Non-blind deconvolution problem When the kernel was known, deconvolution were accomplished based on Richardon-Lucy (RL) or Weiner filtering, which were sensitive to noise and resulted in ringing and artifacts. The aim was to recover the latent silkworm pupa image from the blurred input, with confidence that the restored image could be classified by the intelligent system effortlessly.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
