Application of Interpolation to Multispectral Array Photodetectors

Abstract

The purpose of the research is to apply interpolation to the developed templates of two-layer matrices. Within the framework of this purpose, the following tasks were formulated: research and application of linear and adaptive interpolation methods. The simplest method interpolates each color on its own using some sort of linear interpolation algorithm. This is the fastest method and has the lowest image processing quality. Adaptive methods include various types of interpolation filters: pixel based depending on some heuristic or mathematical models of the local area of the photo around the current point. On the example of three multilayer matrices with cells for registration of the visible and near-IR ranges, the studied interpolation methods were used. Depending on the number of variables used in the equation, using linear interpolation, we get more accurate values when using four known values than when using two known neighboring cells, but the difference in error is small. The standard adaptive method uses 4 adjacent cells to find the desired color. This makes the equation more difficult to calculate. For the same cell, blue and red can only be found using two additional cells. This simplifies the equations by a factor of two, which means that the image processing speed also increases by a factor of 2. At the same time, the error in finding blue and red colors through two additional cells remains approximately the same as when finding green through 4 cells.