Optimal Missing Pixel Estimation Algorithms for Large Detector Arrays
Abstract
The purpose of optimal algorithm design is to develop a cost-effective high-resolution, large-area, digital imaging system, where optimal restoration methods are used to compensate hardware defects, so to reduce over-all system cost. Five modules are implemented in this imaging system: image acquisition, image display, image enhancement, image correction, and image restoration. The focus is on image correction and image restoration. Besides extending the conventional polynomial approximation method to correct both radiometric and geometric distortions, we demonstrate the successful use of the thin-plate spline (TPS) interpolation method for geometric correction since TPS achieves exact mapping. Optimal missing data estimation algorithms including deblurring and denoising are designed to restore images captured from large CCD sensor arrays using butting technique, where 1 to 2 columns of data are missed at the butting edge. We developed the consistency method with separable deblurring, which can deblur the original image and at the same time estimate the missing column(s) exactly, under the condition that no noise is inserted, and the separable blur kernel is exactly known. We also modified the maximum a-posteriori probability (MAP) estimate with the optimization problem solved by mean field annealing (MFA) to fit into this missing data estimation application. It shows more tolerance to perturbations due to noise and inaccurate blur kernel estimation.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 15, 1999
- Accession Number
- ADA379064
Entities
People
- Griff Bilbro
- Hairong Qi
- Wesley E. Snyder
- William Sander
Organizations
- North Carolina State University