CSE

Relevance feedback is an important technique to boost the
retrieval performance in content-based image retrieval (CBIR). There exists
a semantic gap between low-level features and high-level semantic
concepts in CBIR, typical relevance feedback techniques need to perform a
lot of rounds of feedback for achieving satisfactory results. These
procedures are time-consuming and may make the users bored in the
retrieval tasks. In this paper, we propose a novel scheme to study the logbased
relevance feedback technique using the positive and negative
examples for improving retrieval performance and reducing the semantic
gap in CBIR. The proposed system integrates the user’s positive and
negative feedback from all iterations to construct a semantic space to
remember the user’s intent in terms of the high-level semantic features.
The short-term learning further refines the query by updating its
associated weight vector using both positive and negative examples
together with the long-term learning based semantic space. Our proposed
scheme can significantly improve the retrieval performance of semantic
image retrieval for content-based image retrieval.

Image retrieval is an important topic in the field of pattern recognition and artificial intelligence.
Generally, there are three categories of image retrieval methods: text-based, content-based and
semantic-based. In content based image retrieval (CBIR), images are indexed by their visual content
such as color, texture and shape. These low-level image features are insufficient to describe image
contents similar to human visual perception (HVP). In this paper, we proposed a novel approach to
improve optimization of relevance feedback in Content Based Image Retrieval. Firstly, we find
features of an image using the micro-structure descriptor (MSD) are discussed to describe image
features via microstructures. The micro-structure is defined based on the edge orientation and these
low-level features given to Multilayer neural network to find high-level vector generation. Secondly,
we transform low-level features to high-level semantics by means of a multilayer neural network and
these features are employed as the input of a radial basis function network for relevance feedback.
This approach reduces the semantic gap and feature dimensionality.
In this proposed work a new model is presented to optimize the main feature of Relevance
Feedback in the retrieval quality of interactive CBIR. The first idea was to generate micro structured
image (using MSD) to identify low-level features of an image and then characterize images through
neural network, which involves the use of low-level features as support for the high-level vector
generation represented by the neural network. The generated image is used in retrieving the rank
ordered images from the database. Later from the rank ordered images user feedback is given for a
pattern-based search to match the user’s intention. The second idea was to integrate the user's fuzzy
interpretation of image similarity into CBIR system by using Fuzzy Radial Basis Function Network
(FRBFN). This tends to be more flexible than the ordinary relevance feedback algorithms. Finally the
most relevant set of images are retrieved from the database.

This paper provides a mining approach to the research area of relevance feedback (RF) in content-based image retrieval (CBIR). Relevance feedback is a powerful technique in CBIR systems, in order to improve the performance of CBIR effectively. The drawbacks in CBIR are the features of the query image and the semantic gap between low-level features and high level concepts. Especially, Mining Image data is the one of the essential features in this present scenario since image data plays vital role in every aspect of the system such as business for marketing, hospital for surgery, engineering for construction, Web for publication and so on. In this paper, we are proposed an adaptive approach for relevance feedback in CBIR using mining techniques. Where in the processes of feedback we are using a new technique called Image retrieval based on optimum clusters is proposed for improving user interaction with image retrieval systems by fully exploiting the similarity information. The index is created by describing the images according to their color characteristics, with compact feature vectors, that represent typical color distributions.

The purpose of this report is to describe research and solution to the problem of designing a Content Based Image Retrieval, CBIR system. It outlines the problem, the proposed solution, the final solution and the accomplishments achieved. Due to the enormous increase in image database sizes, as well as its vast deployment in various applications, the need for CBIR development arose. Firstly, this report outlines a description of the primitive features of an image; texture, colour, and shape. These features are extracted and used as the basis for a similarity check between images. The algorithms used to calculate the similarity between extracted features, are then explained. Our final result was a MatLab built software application, with an image database, that utilized texture and colour features of the images in the database as the basis of comparison and retrieval. The structure of the final software application is illustrated. Furthermore, the results of its performance are illustrated by a detailed example.
Keywords: Content Based Image Retrieval(CBIR), Similarity, Features and Image Database.

Image retrieval is an important topic in the field of pattern recognition and artificial intelligence. There
are three categories of image retrieval methods: text-based, content-based and semantic-based. In Content based
Image Retrieval (CBIR), images are indexed by their visual content, such as color, texture, shapes. A new image
feature detector and descriptor, namely the micro-structure descriptor (MSD) is discussed to describe image features
via micro-structures. The micro-structure is defined based on the edge orientation similarity, and the MSD is built
based on the underlying colors in microstructures with similar edge orientation. Content-based image retrieval
(CBIR) is the mainstay of image retrieval systems. To be more profitable relevance feedback techniques are
incorporated into CBIR such that more precise results can be obtained by taking user’s feedbacks into account. The
semantic gap between low-level features and high-level concepts handled by the user is one of the main problems in
image retrieval. On the other hand, the relevance feedback has been used on many CBIR systems such as an effective
solution to reduce the semantic gap. The gap is reduced by using the Multitexton Histogram descriptor (MTH) .
In this paper, we propose a novel approach to reduce semantic gap in CBIR using relevance feedback based on the
MSD and MTH descriptors to achieve high efficiency and effectiveness of CBIR in coping with the large-scale image
data. For that reason this paper proposes a method of relevance feedback based on the Microstructure descriptor
(MSD) for efficient feature extraction of an image and Multitexton Histogram descriptor to represents the effective
feature representations. By using this method, high quality of image retrieval on Relevance Feedback can be achieved
in a small number of feedbacks. In terms of efficiency, iteration of feedback is reduced substantially by using the
navigation patterns discovered from the user query log, which reduce the computational processing time.
Keywords--- CBIR, Relevance Feedback, Semantic Gap, Microstructure descriptor (MSD), Multitexton Histogram
descriptor.

Image retrieval is an important topic in the field of pattern recognition and artificial intelligence. There are three categories of image retrieval methods: text-based, content-based and semantic-based. In CBIR, images are indexed by their visual content, such as color, texture, shapes. A new image feature detector and descriptor, namely the micro-structure descriptor [1] (MSD) is discussed to describe image features via micro-structures. The micro-structure is defined based on the edge orientation similarity, and the MSD is built based on the underlying colors in micro-structures with similar edge orientation. Content-based image retrieval (CBIR) is the mainstay of image retrieval systems. To be more profitable relevance feedback techniques are incorporated into CBIR such that more precise results can be obtained by taking user‟s feedbacks into account. The semantic gap between low-level features and high-level concepts handled by the user is one of the main problems in image retrieval. On the other hand, the relevance feedback has been used on many CBIR systems such as an effective solution to reduce the semantic gap. The gap is reduced by using the Multitexton Histogram descriptor [2]. In this paper, a novel framework method called Relevance Feedback is used to achieve high efficiency and effectiveness of CBIR in coping with the large-scale image data. For that reason this paper proposes a method of relevance feedback based on Multitexton Histogram descriptor to represents the effective feature representations, and the Microstructure descriptor (MSD) for efficient feature extraction of an image. By using this method, high quality of image retrieval on Relevance Feedback can be achieved in a small number of feedbacks. In terms of efficiency, iteration of feedback is reduced substantially by using the navigation patterns discovered from the user query log, which reduce the computational processing time.

Content Based Image Retrieval (CBIR) systems attempt to allow users to perform searches in large image repositories. Content-Based Image Retrieval (CBIR) has become one of the most progressive research areas in the past few years. In content Based Image Retrieval, images are retrieved based on color, texture and shape (low level perception). There is a gap between user semantics (high level perception/concepts) and low level perception is called ‘Semantic Gap’. Relevance Feedback (Relevance Feedback) learns association between high level semantics and low level features. While these research efforts establish the basis of CBIR, the usefulness of the proposed approaches is limited. Specifically, these efforts have relatively ignored two distinct characteristics of CBIR systems are semantic gap and human perception of visual content respectively. In this paper, we propose different aspects of the system such as first, we analyze the nature of the Relevance Feedback problem in a continuous representation space in the context of image retrieval. Secondly, a Relevance Feedback based interactive retrieval approach, which effectively takes into account the above two characteristics in CBIR. During the retrieval process, the user's high level query and perception subjectivity are captured by dynamically updated weights based on the user's feedback and finally, the proposed system where user can view/understands the relevance level of the retrieved result of images to his/her given query image. The proposed approach greatly reduces the user's effort of composing a query and captures the user's information need more specifically. We can reduce the user intervention in the CBIR retrieval system.

Image retrieval is an important topic in the field of pattern recognition and artificial intelligence. There are three categories of image retrieval methods: text-based, content-based and semantic-based. In Content-Based Image Retrieval (CBIR), images are indexed by their visual content, such as color, texture, shapes. CBIR has become one of the most active research areas in the past few years. Many visual feature representations have been explored and many systems are built. While these research efforts are established the basis of CBIR, the usefulness of the proposed approaches is limited. Specially, these efforts have relatively ignored two distinct problems of CBIR systems: (1) The semantic gap between high level concepts and low level features; (2) Human perception of visual content. In addition to this, we have the problem of which image analysis models to use in image database to achieve a better CBIR system.
This paper proposes a novel method for combining the user subjectivity in image database and interactive content-based image retrieval (CBIR). It shows a two-step process: 1) Performs image analysis before retrieving an image from the database, which automatically infers which combination of models best are to represents the data of interest to the user and learns continuously during interaction with each user. 2) Effectively takes the above two problems into account in CBIR. In the retrieval process, the user's high level query and perception subjectivity are captured by dynamically updated weights based on the user's feedback. The proposed approach greatly reduces the user's effort of composing a query and captures the user's information.

Image retrieval is an important topic in the field of pattern recognition and artificial
intelligence. There are three categories of image retrieval methods: text-based, content-based
and semantic-based. In Content-Based Image Retrieval (CBIR), images are indexed by their
visual content, such as color, texture, shapes. CBIR has become one of the most active
research areas in the past few years. Many visual feature representations have been explored
and many systems are built. While these research efforts are established the basis of CBIR,
the usefulness of the proposed approaches is limited. Specially, these efforts have relatively
ignored two distinct problems of CBIR systems: The semantic gap between high level
concepts and low level features; Human perception of visual content. In addition to this, we
have the problem of which image analysis models to use in image database to achieve a better
CBIR system.
This paper proposes a novel method for combining the user subjectivity in image
database and interactive content-based image retrieval (CBIR). It shows a two-step process:
Performs image analysis before retrieving an image from the database, which automatically
infers which combination of models best are to represents the data of interest to the user and
learns continuously during interaction with each user. Effectively takes the above two
problems into account in CBIR. In the retrieval process, the user's high level query and
perception subjectivity are captured by dynamically updated weights based on the user's
feedback. The proposed approach greatly reduces the user's effort of composing a query and
captures the user's information.

Image retrieval is an important topic in the field of pattern recognition and artificial intelligence. There are three categories of image retrieval methods: text-based, content-based and semantic-based. In CBIR, images are indexed by their visual content, such as color, texture and shapes. Color and Texture information have been the primitive image descriptors in content based image retrieval systems. Many content-based image retrieval applications suffer from small sample set and high dimensionality problems. Relevance feedback is often used to alleviate those problems. In this paper, an integrating Relevance feedback for content based image retrieval based method is proposed for image mining based on analysis of color Histogram values and texture descriptor of an image and a novel interactive boosting framework to integrate user feedback into boosting scheme and bridge the gap between high-level semantic concept and low-level image features. For this purpose, three functions are used for texture descriptor analysis such as entropy, local range and standard deviation. To extract the color properties of an image, histogram values are used. The combination of the color and texture features of the image provides a robust feature set for image retrieval. Our method has advantage over the classic relevance feedback method in that the classifiers are trained to pay more attention to wrongfully predicted samples in user feedback through a reinforcement training process. It achieves more performance improvement from the relevance feedback than AdaBoost does because human judgment is accumulated iteratively to facilitate learning process.

Content Based Image Retrieval (CBIR) systems attempt to allow users to perform searches in large image repositories. Content-Based Image Retrieval (CBIR) has become one of the most progressive research areas in the past few years. In content Based Image Retrieval, images are retrieved based on color, texture and shape (low level perception). There is a gap between user semantics (high level perception/concepts) and low level perception is called 'Semantic Gap'. Relevance Feedback (Relevance Feedback) learns association between high level semantics and low level features. While these research efforts establish the basis of CBIR, the usefulness of the proposed approaches is limited. Specifically, these efforts have relatively ignored two distinct characteristics of CBIR systems are semantic gap and human perception of visual content respectively. In this paper, we propose different aspects of the system such as first, we analyze the nature of the Relevance Feedback problem in ...

Image retrieval is an important topic in the field of pattern recognition and artificial intelligence. There are three categories of image retrieval methods: text-based, content-based and semantic-based. In Content based Image Retrieval (CBIR), images are indexed by their visual content, such as color, texture, shapes. A new image feature detector and descriptor, namely the micro-structure descriptor (MSD) is discussed to describe image features via micro-structures. The micro-structure is defined based on the edge orientation similarity, and the MSD is built based on the underlying colors in microstructures with similar edge orientation. Content-based image retrieval (CBIR) is the mainstay of image retrieval systems. To be more profitable relevance feedback techniques are incorporated into CBIR such that more precise results can be obtained by taking user’s feedbacks into account. The semantic gap between low-level features and high-level concepts handled by the user is one of the ...

Image retrieval is an important topic in the field of pattern recognition and artificial intelligence. Generally, there are three categories of image retrieval methods: text-based, content-based and semantic-based. In content based image retrieval (CBIR), images are indexed by their visual content such as color, texture and shape. These low-level image features are insufficient to describe image contents similar to human visual perception (HVP). In this paper, we proposed a novel approach to improve optimization of relevance feedback in Content Based Image Retrieval. Firstly, we find features of an image using the micro-structure descriptor (MSD) are discussed to describe image features via microstructures. The micro-structure is defined based on the edge orientation and these low-level features given to Multilayer neural network to find high-level vector generation. Secondly, we transform low-level features to high-level semantics by means of a multilayer neural network and these f...

Image retrieval is an important topic in the field of pattern recognition and artificial intelligence. There are three categories of image retrieval methods: text-based, content-based and semantic-based. In Content-Based Image Retrieval (CBIR), images are indexed by their visual content, such as color, texture, shapes. CBIR has become one of the most active research areas in the past few years. Many visual feature representations have been explored and many systems are built. While these research efforts are established the basis of CBIR, the usefulness of the proposed approaches is limited. Specially, these efforts have relatively ignored two distinct problems of CBIR systems: The semantic gap between high level concepts and low level features; Human perception of visual content. In addition to this, we have the problem of which image analysis models to use in image database to achieve a better CBIR system. This paper proposes a novel method for combining the user subjectivity in i...

Image retrieval is an important topic in the field of pattern recognition and artificial intelligence. There are three categories of image retrieval methods: text-based, content-based and semantic-based. In CBIR, images are indexed by their visual content, such as color, texture and shapes. Color and Texture information have been the primitive image descriptors in content based image retrieval systems. Many content-based image retrieval applications suffer from small sample set and high dimensionality problems. Relevance feedback is often used to alleviate those problems. In this paper, an integrating Relevance feedback for content based image retrieval based method is proposed for image mining based on analysis of color Histogram values and texture descriptor of an image and a novel interactive boosting framework to integrate user feedback into boosting scheme and bridge the gap between high-level semantic concept and low-level image features. For this purpose, three functions are used for texture descriptor analysis such as entropy, local range and standard deviation. To extract the color properties of an image, histogram values are used. The combination of the color and texture features of the image provides a robust feature set for image retrieval. Our method has advantage over the classic relevance feedback method in that the classifiers are trained to pay more attention to wrongfully predicted samples in user feedback through a reinforcement training process. It achieves more performance improvement from the relevance feedback than AdaBoost does because human judgment is accumulated iteratively to facilitate learning process.

"Relevance feedback is an important technique to boost the retrieval performance in content-based image retrieval (CBIR). There exists a semantic gap between low-level features and high-level semantic concepts in CBIR, typical relevance feedback techniques need to perform a lot of rounds of feedback for achieving satisfactory results. These procedures are time-consuming and may make the users bored in the retrieval tasks. In this paper, we propose a novel scheme to study the logbased relevance feedback technique using the positive and negative examples for improving retrieval performance and reducing the semantic gap in CBIR. The proposed system integrates the user’s positive and negative feedback from all iterations to construct a semantic space to remember the user’s intent in terms of the high-level semantic features. The short-term learning further refines the query by updating its associated weight vector using both positive and negative examples together with the long-term learning based semantic space. Our proposed scheme can significantly improve the retrieval performance of semantic image retrieval for content-based image retrieval."

Image retrieval is an important topic in the field of pattern recognition and artificial intelligence. There are three categories of image retrieval methods: text-based, content-based and semantic-based. In Content-Based Image Retrieval (CBIR), images are indexed by their visual content, such as color, texture, shapes. CBIR has become one of the most active research areas in the past few years. Many visual feature representations have been explored and many systems are built. While these research efforts are established the basis of CBIR, the usefulness of the proposed approaches is limited. Specially, these efforts have relatively ignored two distinct problems of CBIR systems: (1) The semantic gap between high level concepts and low level features; (2) Human perception of visual content. In addition to this, we have the problem of which image analysis models to use in image database to achieve a better CBIR system. This paper proposes a novel method for combining the user subjectiv...

The purpose of this report is to describe our research and solution to the problem of designing a Content Based Image Retrieval, CBIR system. It outlines the problem, the proposed solution, the final solution and the accomplishments achieved. Due to the ...