An Approach for Detection of the Components in Brain MRI Using Vector Quantization and Morphological Operations based Segmentation

Image segmentation is a technique in image processing where an image is divided into meaningful structures to simplify its representation. Vector quantization helps to map the continuous pixel of input space to discrete pixels in the output image. Here the advantage is to minimize the information loss. The approach in this paper focuses on the spatial as well as the gray level value of the image to effectively derive beneficiary result. The detection of the important components in brain MRI image has become a challenging task in terms of the performance of different existing algorithms. Here the proposed method involves preprocessing of the brain image with tumor, vector quantization, adaptive binarization, application of morphological operations. The calculation of the tumor area is done using the proposed algorithm and manually using active contour method where the accuracy of the proposed method is calculated. In this paper the method extracts the main components of the brain with tumor which includes the skull, the tumor, the gray matter and the white matter. The method could easily extract the tumor of any size from a brain image with tumor having more accuracy than any other existing methods. Keywords— Image segmentation, Vector Quantization, Adaptive binarization, Morphological operations, MRI images, Contour method, preprocessing.