Bad pixel replacement based on spatial statistics for IR sensor

Abstract

IR focal plane arrays typically contain bad pixels. Bad pixels have to be corrected because those can significantly impair the performance of target-detection algorithms. On the other hand, particularly as an aid to visual interpretation, it is desirable to replace the bad pixels. IR image contains spatial information and is correlative in spatial domain. In spatial statistics the semivariogram is an important function that relates semivariance to sampling lag. This function can characterize the spatial dependence of each point on its neighbor and provide a concise and unbiased description of the scale and pattern of spatial variability. One of the main reasons for deriving semivariogram is to use it in the process of estimation. Kriging is an interpolation and estimation technique that considers both the distance and the degree of variation between known data points when estimating values in unknown areas. In this paper a new technique based on spatial statistics is developed for bad pixel replacement. The main objective of the technique is to replace bad pixels through Kriging estimation. Theory analysis and experiments show that the method is reasonable and efficient.

Identification and replacement of defective pixel based on Matlab for IR sensor

“…To achieve this goal, we propose a novel identification method which combines the median filtering algorithm [9] with the improved standard deviation algorithm. The proposed method is significantly different from traditional methods.…”