Comparative Analysis of Spatial Filters for Image Rectification and Computational Complexity

Abstract

Image Rectification is an important preprocessing step in many fields such as computer vision, pattern matching and many real time image-processing applications. Rectifying different things is image Rectification. But here the algorithm that helps to rectify impurity presents in image is proposed. This impurity can be noise and distortions. The sources of impurity in digital images arise during image acquisition (digitization) and /or transmission. Impurity has two kinds of properties 1) Spatial properties 2) Frequency properties. Frequency properties refer to the frequency content of impurity in the Fourier sense. The spatial characteristics are concerned with statistical behavior of the gray level values in the impurity component. So the proposed algorithm focuses in spatial properties of impurity. There are two types of spatial filters. But noise reduction can be achieved effectively with a nonlinear filter whose basic function is to compute the median graylevel value in the neighborhood in which the filter is located. There exist many works on minimizing impurity of images by using spatial filters. The most widely used spatial filters are Median Filter, Kuwahara filter and Gaussian filter.

All are responsible to minimize impurity to some extent. The proposed algorithm that

helps to minimize impurity is better than these filters mentioned above. Here, checking of the minimization of error is done by calculating RMS errors between input image and output image. The proposed algorithm is also based on the spatial filters. Experiment

showed that the final rectified images are satisfactory and it makes it easy for further image manipulation.

The efficiency of an algorithm depends on the complexity (time and space). Time

complexity is total time required for execution, whreas space complexity is total memory

space required for execution of an algorithm. Complexity of Median Filter, Kuwahara filter and Gaussian filter and proposed filter are analysed for better efficiency.

Keywords: