Vehicle detection and tracking plays an effective significant role in the area of traffic surveillance system where efficient management safety is main concern. In this paper, we discuss address issue detecting vehicle / data from video frames. Although various researches have been done many methods implemented, still has room for improvements. With a view to do improvements, it proposed develop unique algorithm recognition using Gaussian mixture model blob methods. First, differentiate...

In the past decade, various new and impressive applications have been developed implemented on drones, for instance search rescue, surveillance, traffic monitoring, weather monitoring so on. The current advances in drone technology provoked significant changes enabling drones to perform a wide range of missions with increasing level complexity. Missions such as rescue or forest surveillance require large camera coverage thus making suitable tool advanced tasks. Meanwhile, trend deep learning...

Deep learning architectures particularly Convolutional Neural Network (CNN) have shown an intrinsic ability to automatically extract the high level representations from big data. CNN has produced impressive results in natural image classification, but there is a major hurdle their deployment medical domain because of relatively lack training data as compared general imaging benchmarks such ImageNet. In this paper we present comparative evaluation three milestone i.e. LeNet, AlexNet and...

Modality corresponding to medical images is a vital filter in image retrieval systems, as radiologists or physicians are interested only one of radiology e.g CT scan, MRI, X-ray. Various handcrafted feature schemes have been proposed for modality classification. On the other hand not enough attempts made deep learned extraction. A comparative evaluation both and features classification presented this paper. The experiments performed on IMAGECLEF 2012 data. After carrying out it shown that...

Drones play a pivotal role in various industries of Industry 4.0. For achieving the application drones dynamic environment, finding clear path for their autonomous flight requires more research. This paper addresses problem navigation an drone based on visual scene information. A deep learning-based object detection approach can localize obstacles detected scene. Considering this approach, we propose solution framework that includes masking with color-based segmentation method to identify...

Hoaxes are non-malicious viruses. They live on deceiving human’s perception by conveying false claims as truth. Throughout history, hoaxes have actually able to influence a lot of people the extent tarnishing victim’s image and credibility. Moreover, wrong misleading information has always been distortion growth. Some were created in way that they can even obtain personal data convincing victims those required for official purposes. different from spams masquerade themselves through address...

Modality corresponding to medical images is a vital filter in image retrieval systems. This article presents the classification of modalities based on usage principles hyper-dimensional computing and reservoir computing. It demonstrated that highest accuracy proposed method par with best classical for given dataset (83% vs. 84%). The major positive property it does not require any optimization routine during training phase naturally allows incremental learning upon availability new data.

Financial data are characterized by non-linearity, noise, volatility and chaotic in nature thus making the process of forecasting cumbersome. The main aim forecasters is to develop an approach that focuses on increasing profit being able forecast future stock prices based current data. This paper presents empirical long term financial using Parallel non-linear auto-regressive with exogenous input (P-NARX) network trained Bayesian regulation algorithm. experimental results mean absolute...

Metocean time-series data is generally classified as highly chaotic thus making the analysis process tedious. The main aim of forecasting to obtain an effective solution for offshore engineering projects, such environmental conditions vital choices made during planning and operational stage which must be efficient robust. This paper presents empirical using a hybrid neural network model by performing multi-step-ahead forecasts. proposed trained gauss approximated Bayesian regulation...

Reservoir simulation provides information about the behaviour of a reservoir in various production and injection conditions. simulator is used to predict future performance field. However, heterogeneity uncertainty field cause some obstacles selecting best calculation oil, water gas components that lead system oil gas. Due intrinsic models, large number computational resources such as runs long processing time are required properties reservoir. This paper presents an application Surrogate...

This paper presents the Latent Semantic Analysis (LSA) Model in Automatic Text Summarization (ATS) on single English document mobile Android platform. Readers are drowned information while starved of knowledge. Millions articles uploaded into website every day. Quite often, lengthy text presented online but shorter summarized texts preferred by readers. There exists research gap as most extractive summarizations based syntactic appearance words. Thus, objective this is to investigate LSA...

The increasing number of medical images various imaging modalities is challenging the accuracy and efficiency radiologists. In order to retrieve from databases, radiologists will confine their search image modality. this paper, we present an improved feature represent for modality classification. proposed descriptor ensemble that combines Harris Corner encoded by SIFT algorithm fused with Local Binary Pattern. Furthermore, propose classifier surrogate splits be used in classification improve...

This paper presents a Droopy Mouth Detection Model in stroke warning. The objective of this is to take up the challenge provide early detection through mouth drooping mobile Android platform. To achieve that, specialized library, Google Mobile Vision utilized detect facial landmark such as corners and obtain coordinates landmarks or key points for further processing. inputs proposed droopy model were taken from Web Images National Cheng Kung University (NCKU) Robotics Face datasets. system...

Hoaxes are non malicious viruses. They live on deceiving human?s perception by conveying false claims as truth. Throughout history, hoaxes have actually able to influence a lot of people the extent tarnishing victim?s image and credibility. Moreover, wrong misleading information has always been distortion growth. Some were created in way that they can even obtain personal data convincing victims those required for official purposes. different from spams masquerade themselves through address...

A novel approach to robust and flexible person tracking using an algorithm that integrates state of the arts techniques; Enhanced Person Detector (EPD) Kalman filtering algorithm. This proposed employs multiple instances Filter with complex assignment constraints Graphics Processing Unit (GPU-NVDIA CUDA) as a parallel computing environment for persons even in presence occlusion. filter is recursive which predict variables further uses observed data correct predicted value. Data association...

In the area of traffic flow monitoring, planning and controlling, a video based detection tracking plays an effective significant role where management safety is main concern. The goal project to recognize moving vehicles track them throughout their life spans. this paper, we discuss address issue detecting vehicle/traffic data from frames with increased real time processing. Although various researches have been done in many methods implemented, still has room for improvements. With view do...

Novelty detection is an important functionality that has found many applications in information retrieval and processing. In this paper we propose a novel framework deals with novelty for multiple-scene image sets. Working wildlife data, the starts segmentation, followed by feature extraction classification of blocks extracted from segments. The labelled are then scanned through to generate co-occurrence matrix object labels, representing semantic context within scene. matrices undergo...