Modified Alignment Technique Using Matched Important Features

In image processing the exact alignment of the two images takes great and wide attention, and achieving the perfect results is very important because of the importance of that process in many scientific applications. In this work, a strategy has been presented that blends pixel-based alignment techniques with those that focus on the important features or characteristics of the image, and the work on this technique has been tested and proven to be effective in terms of high accuracy and fast delivery. The working principle of the proposed method is divided into four non-overlapping phases: the important features extraction, the discovery of the matching features, the creation of the perfect conversion coefficients for the image, the implementation of the alignment process The extraction step of the features is considered important or essential ,which greatly improves the alignmentprocedure. The fundamental step is the extraction of the main key points in each image with the use of a SIFT detector. The matching features between the two images are then searched using a method based on the principle of grouping adjacent pixels for each important feature. The alignment process uses the Iterative Closest Point algorithm to perform a rigid transform with the use of the matched key points. The results show that the alignment process using the suggested method has yielded quite efficient results and it is highly important finding the interesting features and the strong correspondence point