Content Based Image Retrieval

SKU: PAN_IPM_038 Categories: ,
Ask For Price



In recent years, instance-level-image retrieval has attracted massive attention. Several researchers proposed that the representations learned by convolution neural network (CNN) can be used for image retrieval tasks. In this study, the authors propose an effective feature encoder to extract robust information from CNN. It consists of two main steps: the embedding step and the aggregation step. Moreover, they apply the multi-task loss function to train their model in order to make the training process more effective. Finally, this study proposes a novel representation policy that encodes feature vectors extracted from different layers to capture both local patterns and semantic concepts from deep CNN. Do they call this? multi-level-image representation? which could further improve the performance. The proposed model is helpful to improve retrieval performance. For the sake of comprehensively evaluating the performance of their approach, they conducted ablation experiments with various convolutional NN architectures. Furthermore, do they apply their approach to a concrete challenge? Alibaba large-scale search challenge. The results show that their model is effective and competitive.


Imitations are not new to humanity however as an opportunity is very old trouble. In the beyond it changed into restrained craftsmanship and writing but did now not have an effect on the overall population. These days, due to the headway of automatic image handling software programs and changing gadgets, an image can be results managed and changed. It is alternatively tough for people to apprehend outwardly whether or now not or now not the photo is unique or manipulated. There is a speedy increment in digitally controlled falsifications in great media and on the Internet. This sample shows authentic vulnerabilities and abatements of the credibility of digital photographs. In this manner, growing techniques to check the honesty and realness of the superior snapshots is essential, especially considering that the picture is delivered as evidence in a courtroom docket of regulation, as facts things, as part of restorative records, or as coins related opinions. Image retrieval is one of the famous and wanted techniques nowadays. Here the dataset is trained into our code and the input test image gives a retrieved all images at the output by means of a neural network.

System Analysis

Existing Systems

  • SIFT based sparse features representation
  • GLCM and Local Binary Pattern analysis


  • Sparse representation doesn’t contain information about texture.
  • It has poor discriminatory power
  • It is poor to characterize the contrast information

Proposed Method

Robust object recognition for Content-based image retrieval on multimedia images based on, Discriminative robust Local binary pattern and Local ternary Pattern analysis


  • Better efficiency and less sensitive to noise
  • Better accuracy
  • Processing time is less

Block diagram:


Content Based Image Retrieval

HSV Colour Mode

The HSV stands for the Hue, Saturation, and worth supporting the artists (Tint, Shade, and Tone). The frame of reference is an extremely hexagon in Figure a combine. (a).

And Figure a combines of.(b) a browse of the HSV color model. the price represents the intensity of a color, that’s decoupled from the color information inside the drawn image.

The hue and saturation components area unit intimately related to the means the human eye perceives color resulting in image method algorithms with a physiological basis.

As hue varies from zero to one.0, the corresponding colors vary from red, through yellow, green, cyan, blue, and magenta, back to red, therefore there are actually red values every at zero and one.0.

As saturation varies from zero to one.0, the corresponding colors (hues) vary from unsaturated (shades of gray) to fully saturated (no white component). As a value, or brightness, varies from zero to one.0, the corresponding colors become brighter and brighter.

DRLBP Descriptor:

The descriptor local binary pattern is used to compare all the pixels including the center pixel with the neighboring pixels in the kernel to improve the robustness against the illumination variation. An LBP code for a neighborhood was produced by multiplying the threshold values with weights given to the corresponding pixels and summing up the result. LBP codes are weighed using a gradient vector to generate the histogram of robust LBP and discriminative features are determined from the robust local binary pattern codes. DRLBP is represented in terms of a set of normalized histogram bins as local texture features. It is used to discriminate the local edge texture of the face invariant to changes in contrast and shape.?

Types of Neural Networks:

Artificial Neural Network

  • Backpropagation networks
  • General Regression Neural Networks

DTREG implements the most widely used types of neural networks:

  1. a) Multilayer Perceptron Networks (also known as multilayer feed-forward networks),
  2. b) Cascade Correlation Neural Networks,
  3. c) Backpropagation networks (BPN)
  4. d) General Regression Neural Networks (GRNN)

This network has an input layer (on the left) with three neurons, one hidden layer (in the middle) with three neurons, and an output layer (on the right) with three neurons.

There is one neuron in the input layer for each predictor variable. In the case of categorical variables, N-1 neurons are used to represent the N categories of the variable.

Hardware Requirements

  • system
  • 4 GB of RAM
  • 500 GB of Hard disk



[1] Settles, B., Craven, M., Ray, S.:?Multiple-instance active learning?. Proc. Advances in Neural Information Processing Systems, Vancouver, Canada, 2007, pp. 1289? 1296

[2] Gordo, A., et al.:?Deep image retrieval: learning global representations for image search?. Proc. European Conf. on Computer Vision, Amsterdam, The Netherlands, October 2016, pp. 241? 257

[3] Lowe, D.G.:?Distinctive image features from scale-invariant keypoints?, Int. J. Comput. Vis., 2004, 60, (2), pp. 91? 110

[4] Bay, H., Tuytelaars, T., Van Gool, L.: ?Surf: speeded up robust features?. Proc. European Conf. on Computer Vision, Graz, Austria, May 2006, pp. 404? 417

[5] Qiu, G.: ?Indexing chromatic and achromatic patterns for content-based color image retrieval?, Pattern Recognit., 2002, 35, (8), pp. 1675? 1686

Customer Reviews

There are no reviews yet.

Be the first to review “Content Based Image Retrieval”

Your email address will not be published. Required fields are marked *