Search

In magnetic resonance (MR) brain picture investigation, brain picture segmentation (BPS) is usually utilized for identifying, estimating and investigating the inner structure of the brain and also the neural structure of the brain. BPS helps the specialists to concentrate on specific regions inside the brain and to examine them. However, BPS is a difficult task because of high similitudes and connections of force among various districts of the brain picture. Moreover, due to unoptimized supervised technique and usage of more data in representing the image, accuracy gets affected and consumes more time. So there is a need to develop an algorithm which provides high accuracy in segmenting regions faster, even though the similarities between the tissues are more in brain. This framework uses the Local Binary Pattern, modified histogram of oriented gradients (MHOG), local average based kinetic gas molecular optimization (LA-KGMO) and modified self- organizing map (MSOM) for efficient segmentation of the MR brain regions. The feature depth of the extracted texture feature is increased by utilizing the modified histogram of oriented gradients (MHOG) whose are extracted and cascaded. LA-KGMO is used to optimize the extracted MHOG features. In order to avoid random generation of weight vector as in case of simple SOM, a fixed design unsupervised MSOM is used. By optimizing the features used in the training of MSOM, dimension gets reduced. As all the approaches are combined in this proposed method, this will increase the performance of the system. The obtained results are compared with other methodologies such as Convolutional neural network, artificial neural network, K-means integrated fuzzy C-means and support vector machine. This method segment the whole brain MR image into tumor region, White Matter, Gray Matter and Cerebrospinal Fluid. To analyze the performance of the proposed method, it is compared with existing methods in terms of accuracy, RMSE, JAC and HD and produced superior performance over the conventional methods. The proposed architecture is tested with the Whole Brain Atlas dataset. In this work, LA-KGMO Optimized MR brain image clustering using MSOM is proposed and implemented. The analysis results show that the proposed technique attains an accuracy of 0.97 average for around 500 images from Whole Brain Atlas database. The computation time of this proposed method is 4 s. [ABSTRACT FROM AUTHOR]

Based on the rapid increase in the demand for people counting and tracking systems for surveillance applications, there is a critical need for more accurate, efficient, and reliable systems. The main goal of this study was to develop an accurate, sustainable, and efficient system that is capable of error-free counting and tracking in public places. The major objective of this research is to develop a system that can perform well in different orientations, different densities, and different backgrounds. We propose an accurate and novel approach consisting of preprocessing, object detection, people verification, particle flow, feature extraction, self-organizing map (SOM) based clustering, people counting, and people tracking. Initially, filters are applied to preprocess images and detect objects. Next, random particles are distributed, and features are extracted. Subsequently, particle flows are clustered using a self-organizing map, and people counting and tracking are performed based on motion trajectories. Experimental results on the PETS-2009 dataset reveal an accuracy of 86.9% for people counting and 87.5% for people tracking, while experimental results on the TUD-Pedestrian dataset yield 94.2% accuracy for people counting and 94.5% for people tracking. The proposed system is a useful tool for medium-density crowds and can play a vital role in people counting and tracking applications. [ABSTRACT FROM AUTHOR]