An efficient parameterless quadrilateral-based image segmentation method

A New Method for Colour Image Segmentation

“…The evaluation is performed comparing the obtained results by means of the proposed algorithm with results obtained by other three methods: HCI fuzzy segmentation [11], segmentation using situational DCT (Discrete Cosine Transform) descriptors [12] and parameterless quadrilateral-based image segmentation method [10]. Figures 7,8,and 9 show image segmentation using the different methods.…”

“…In general, the split and merge method begins with an initial and no homogeneous image partition, then keeps on splitting it until homogeneous partitions are obtained. A common data structure used in the implementation of this procedure is the quad tree representation [10]. After a division Copyright © 2006Copyright © -2016 by CCC Publications step, usually, many small and fragmented regions appear connected in some way during the merging phase.…”

“…The effectiveness of the proposed method is compared with a parameter less quadrilateral-based image segmentation method [10] with a fuzzy segmentation HCI [11], and with a image segmentation using situational descriptors DCT [12]. The proposed algorithm performs the best segmentation and requires the least processing time compared with other methods used in its evaluation.…”

Image Segmentation Using Information Bottleneck Method

“…The principal approaches in the second category are based on partitioning an image into regions that are similar according to a set of predefined criteria. Thresholding, region growing, histogram clustering, split-and-merge, and random fields are examples of methods of this category [1], [2], [10], [12], [13], [18], [30]. For our purposes, we briefly review the thresholding, histogram clustering, and split-and-merge algorithms.…”

Brain Tumor Detection Using Artificial Neural Network Fuzzy Inference System (ANFIS)

“…It is a reliable tool for the glioma tumor series. The EEG in vascular lesions shows abnormality on first instance where as a CT scan shows abnormal on the third or fourth day .Medical Resonance images include a noise which is created due to operator's method of detection which can lead to serious inaccuracies in classification of brain tumor [1]. With increasing problems of brain, it is vital to develop a system with novel algorithms to detect brain tumor efficiently .The present method detects tumor area by darkening the tumor portion and enhances the image for detection of other brain diseases in human being.…”