Colour image annotation using hybrid intelligent techniques for image retrieval
Summary (2 min read)
Introduction
- Content-based image retrieval (CBIR) is a technique that involves retrieving specific images from image databases primarily based on features that could be automatically derived [1].
- Section 2 provides a literature review of CBIR systems.
- Section 3 describes the proposed CBIR technique in a novel MapReduce neural network framework for large image databases.
A. What is MapReduce?
- MapReduce is a distributed computing framework to support parallel computations over large datasets in multiple petabytes of storage available on clusters of computers.
- The concept originates from map and reduce functions commonly used in functional programming like Lisp, and have been improvised in MapReduce framework, which transform a list of pairs <key, value> into a list of values.
- These partitioned input sets can be processed in parallel on different machines.
- The MapReduce framework consists of a single master Job Tracker and one slave Task Tracker per cluster node.
- Figure 1 shows an example of high level architecture of the MapReduce Framework implemented in Hadoop with a cluster setup consisting of 1 master node and 2 slave nodes.
B. Neural Network Ensembles for CBIR
- The shortcomings and problems encountered with traditional methods of image retrieval have led to the rise of interest in CBIR techniques.
- To address these limitations, the authors have adopted a hybrid technique for CBIR based on fuzzy logic and neural networks within a MapReduce distributed framework for processing very large image collections in the cloud.
- The authors combine the colour features such as Red, Orange, Yellow, Green, Cyan, Blue, Purple, Magenta, Pink, Black, White and Grey, as well as fuzzy terms of colour content such as 'no colour', 'very low', 'low', 'medium', 'high', and 'very high' for colour image classification and perform retrieval using neural networks.
- This involves steps which include fuzzy interpretation of user queries, neural network to train the queries and a technique for the fusion of multiple queries.
- The advantage of NNE is that different networks such as multilayer perceptrons, radial basis functions neural networks, and probabilistic neural networks, can take as inputs samples characterised by different feature of colour and content type.
C. MapReduce Neural Network Implementation
- The parallel framework offered by MapReduce is highly suitable for the neural network ensemble technique of their proposed CBIR framework.
- MapReduce uses two functions called Map and Reduce that process list of pairs <key, value>.
- The neural network ensemble is applied into a two-stage MapReduce process, with one for training the classifier and the other for validating the classifier.
- The user query is used to match with the images in the data set using neural networks for the classification and indexed based on percentage of colour relevance.
- The classified images are retrieved and are displayed in descending order of colour relevance.
Did you find this useful? Give us your feedback
Citations
References
6,447 citations
"Colour image annotation using hybri..." refers background in this paper
...Figure 2 shows an example of extracting and storing colour features in database....
[...]
...One of the important challenges of Content-based Image Retrieval (CBIR) is minimize the semantic gap between low level features (colour, texture, shape) and high level semantic concepts (car, tree, building) described in an image [1]....
[...]
3,433 citations
"Colour image annotation using hybri..." refers background or methods in this paper
...Some of the techniques used for reducing the semantic gap have been developed and presented in various research papers [2][3]....
[...]
...Many image retrieval systems and automatic image annotation techniques are surveyed in [3] along with critical challenges involved in deploying current image retrieval techniques for real world applications....
[...]
2,117 citations
1,765 citations
"Colour image annotation using hybri..." refers background in this paper
...Many of them realize another problem which is dependency on the training dataset to learn the models [9]....
[...]
1,475 citations
"Colour image annotation using hybri..." refers methods in this paper
...Some of the techniques used for reducing the semantic gap have been developed and presented in various research papers [2][3]....
[...]
Related Papers (5)
Frequently Asked Questions (13)
Q2. Why is it important to retrieve the images at various locations?
Due to adavances in technology such as cloud computing, it is very important to retrieve the imageseffectively and efficiently stored at various locations.
Q3. What classes were used for each image?
The authors used the following fuzzy classes for colour contents for each image: very low [0.05, 0.1], low [0.11, 0.35], medium [0.36, 0.65], high [0.66, 0.80], very high [0.81, 1.0].
Q4. What is the main purpose of the proposed framework?
The most important step of training the classes for the fusion of queries facilitates in accuracy of results and weadopt a neural network based technique for this purpose.
Q5. What is the concept of map and reduce?
The Map functions are invoked in a distributed environment across multiple machines by automatically partitioning the input data into a set of K splits.
Q6. How did the authors calculate the colour content of the images?
The authors calculated the colour content by incrementing the relevant colour counter when the relevant colour pixel was found and the total for the entire colours were also calculated to estimate the percentage of each colour.
Q7. What is the main purpose of the CBIR?
It takes care of balancing the tasks and optimising the overall runtime, making their proposed CBIR real-time efficient for processing very large image datasets.
Q8. What is the advantage of using a parallel framework for CBIR?
The parallel framework offered by MapReduce is highly suitable for the neural network ensemble technique of their proposed CBIR framework.
Q9. What is the role of the master in the MapReduce framework?
The master is responsible for scheduling the jobs' component tasks on the slaves, monitoring them and re-executing the64 2012 12th International Conference on Hybrid Intelligent Systems (HIS)failed tasks..
Q10. What is the advantage of using a hybrid technique for CBIR?
In situations that require processing of large image datasets, NNE is produces more accurate outcomes as they are robust and efficient than single neural network.
Q11. What is the main idea of the paper?
Though parallel processing would reduce the speed to a great extent, the additional overhead of pooling results from the distributed clusters has to be considered.
Q12. What is the purpose of the map and reduce framework?
This would facilitate batch processing of the files and features during MapReduce search tasks that are assigned by the job tracker to the task tracker present in each of the nodes to facilitate parallel processing in a cluster environment.
Q13. What is the purpose of this paper?
This paper evaluates various CBIR systems developed using conventional as well as computational intelligence techniques and proposes a novel MapReduce Neural Network framework for CBIR in five stages.