Case Study : Ontology-Based Medical Image Understanding

Data mining has popular technology for extracting interesting information for multimedia data sets, such as audio, video, images, graphics, speech, text and combination of several types of data set. Multimedia data are unstructured data or semi-structured data. These data are stored in multimedia database, multimedia mining which is used to find information from large multimedia database system, using multimedia techniques and powerful tools. The current approaches and techniques are explained for mining multimedia data. This paper analyzes about the use of essential characteristics of multimedia data mining, retrieving information is one of the goals of data mining and different issues have been discussed.

Thus our proposed optimal feature subset selection based on multi-level feature subset selection produced better results based on number of subset feature produced and classifier performance. The future scope of the work is to use these features to annotate the image regions, so that the image retrieval system can retrieve relevant images based on image semantics.

Abstract - Content based image retrieval is an image search technique which uses content of features such as colour, texture, shape, etc. to find relevant images from large collection of data according to user’s request query image on the basis of similarity in features or content that means feature from query image compared with features from image database. The problem arises that large number of features may cause high dimensionality curse. To avoid dimensionality curse, feature selection method is used

In user studies, clinicians have indicated that modality is one of the most important filters that they would like to be able to limit their search by [2]. Due to the immense need for effective and accurate medical image classification, new trends for classification of medical images have been investigated for the past few years [3]. Textual annotation is an approach by which text can be attached, but again, this is a time consuming process and requires human intervention. By targeting visual features of the medical image classification of the modality can be done, which promises a better solution to this

Human interpreters can combine data from different sensors by either observing multiple data alternatively or merging multiple data into a single visual. An example of the fused image would be color fusion. It is effective in separating distinguishable features, for example, road and grass, thus it is helpful in detecting objects. However, it is not suitable for images in which the contents are similar, for example, trees and grass. There are different opinion and preferences between images by single sensor and images from multiple sensors. Generally speaking, fused images can help human interpreters to perceive data more effectively. However, they are also problematic in that data must be stretched in order to cohere to one another in the process of merging. Thus, distortion occurs which ruins the aesthetics of the visual representation. This might be a factor which hinders interpreters’ preference for fused images. Therefore, in the fusion process, we must consider human interpreters and make sure the fused images are helpful in perceiving the data.

2It should be known that a single memory is a complex construction. For example, when people think of a simple object the brain retrieves the objects name, its shape, function, and anything else that closely ties itself to that object. 2Each part of this memory is from a different region of the brain. 3(Mohs and Turkington) As stated by April Holladay, “memories of concepts and ideas are related to sense experiences because we extract the essence from sensed experiences to form generalized concepts.” And in the past, many experts explained memory as a ‘filing cabinet’ full of memory folders where

(It can not be considered as if it were isolated in this small part of the process which concerns the retina of the eye.) One sees only what one looks at. To look is to choose. And it is by this choice that what we see appears in our field without being nevertheless within reach. It is a touch that places us in relation to things. (Take a few steps with your eyes closed and you will realize how touching is to see, although in a less mobile and less extensive way.) It is never on a single object that our gaze is focused but on The relationship between us and things. Vision is an incessant activity, sweeping the world around us in a circle that defines our microcosm. "

The term ontology is derived from the Greek word, with “onto” meaning “being”, and “logos” usually interpreted as “science”; so that ontology, as customarily understood, is the science or study of being (Lawson, 2004). Also, ontology is a branch of metaphysics concerned with the nature and relations of

Images with higher similarity than the threshold are returned as an output. System which uses the content of the object for Information Retrieval is the CBIR.

As technology begins to embody human-like characteristics via artificial intelligence, it is beneficial to keep in mind that artificial intelligence is also present in search algorithms on web browsers. Nowadays, artificial intelligence is found in the basic functions of computers and smartphones, a necessity of the twenty-first century. The most common type of artificial intelligence today is known as narrow AI, which is created to perform a narrow task, such as driving a car or recognizing a face. The goal of many researchers is general AI, which would

In this step, the proposed ontologies will be used to guide information extraction. The databases listed in Table 3 may be used as data sources to elicit the faults of

Multimedia data mining is used for extracting interesting information for multimedia data sets, such as audio, video, images, graphics, speech, text and combination of several types of data set. Fig.1 explains about multimedia mining it is a subfield of data mining which is used to find interesting information of implicit knowledge. Multimedia data are classified into five types, there are (i) text data (ii) image data (iii) audio data (iv)video data and (v) electronic and digital ink [2]. Text data

Abstract—Present work targets the detection of humans in images and videos. Our focus is on developing robust feature extraction algorithms that encode image regions as high dimensional feature vectors that support high accuracy human/non-human decisions. To test our feature sets we adopt a relatively simple learning framework that uses linear Support Vector Machines to classify each possible image region as a Human or as a non-Human. This work makes three main contributions. Firstly, we introduce grids of locally normalized Histograms of Oriented Gradients (HOG) as descriptors for people detection in static images. The HOG descriptors are computed over dense and overlapping grids of spatial blocks, with image gradient orientation features extracted at fixed resolution and gathered into a high dimensional feature vector. They are designed to be robust to small changes in image contour locations and directions, and significant changes in image illumination and color. Secondly, the Human detection algorithm has to be executed on modern parallel platforms to achieve the detection goal in real time. Thirdly, alert video monitoring station with an indication. It enhance the Real Time Streaming Protocol (RTSP) based Surveillance cameras purpose.

Although tremendous progress has been made in past decade to detect the unattended objects especially in public places namely railway stations, shopping malls etc but still fails to meet the desired requirement. An automated mechanism with robust approach for detecting objects left unattended in consumer world resulted in successful commercialization. The successful detection of object detection mainly relies on object extraction and objects classification and based on this approaches a novel algorithm namely intelligent vision based analyzer for semantic analysis is proposed for unattended object analyzer. Finally experimental results show good efficiency and better performance over near competitors for real-time consumer video surveillance system for public safety application in big cities.

Data mining has proved to popular for extracting interesting information for multimedia data sets, such as audio, video, images, graphics, speech, text and combination of several types of data set. Multimedia data are unstructured data or semi-structured data. These data are stored in multimedia database, multimedia mining find information from large multimedia database system, using multimedia techniques and powerful tools.