Unlock instant, AI-driven research and patent intelligence for your innovation.

Image retrieving method based on salient region

An image retrieval and region technology, applied in neural learning methods, special data processing applications, instruments, etc., can solve the problem that the retrieval accuracy needs to be improved, and achieve the effect of improving accuracy

Active Publication Date: 2017-10-10
成都星亿年智慧科技有限公司
View PDF5 Cites 16 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] However, these methods mine the attributes of the image from the perspective of recognizing the image, rather than mining the attributes of the image from the perspective of understanding the image, and the retrieval accuracy still needs to be improved.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Image retrieving method based on salient region
  • Image retrieving method based on salient region
  • Image retrieving method based on salient region

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0046] For the convenience of description, the relevant technical terms appearing in the specific implementation are explained first:

[0047] CNN (Convolutional Neural Network): convolutional neural network;

[0048] RPN (Region Proposal Network): regional positioning network;

[0049] LSTM (Long Short Time Memory): long short-term memory network;

[0050] ROI (Region of Interest): Region of interest;

[0051] FC (Fully Connect): full connection; MPoC (Max Pooling of Convolutional features): maximum pooling convolution features;

[0052] SPoC (Sum Pooling of Convolutional features): and pooling convolutional features.

[0053] figure 1 It is a flow chart of the image retrieval method based on the salient region of the present invention.

[0054] In this example, if figure 1 Shown, the present invention a kind of image retrieval method based on salient region, comprises the following steps:

[0055] S1. Extract the salient region of the input image

[0056] S1.1, take an...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an image retrieving method based on a salient region. The image retrieving method includes the steps of retrieving the salient region of an image, describing and conducting pooling encoding and other treatment on the salient region, extracting local CNN features of the image and global CNN features of the image, and then through the global and local CNN features of the image to be retrieved, conducting same-category retrieval and same-object retrieval on an imagery database. According to the image retrieval method based on the salient region, the accuracy of image retrieval is improved.

Description

technical field [0001] The invention belongs to the technical field of image retrieval, and more specifically relates to an image retrieval method based on a salient region. Background technique [0002] Content-based image retrieval technology (CBIR, content-based image retrieval) means that the object used for searching is an image itself, or a feature description of the image content. Most existing methods use the underlying visual features of images, such as sift descriptors, and use bag-of-words (BoW), Fisher vectors (FV) or vector locally aggregated descriptors (VLAD) to encode sift descriptors. But the performance of most traditional image retrieval algorithms can't meet people's requirements. The reason is mainly the difference between the semantic understanding of low-level features and high-level features, that is, the semantic gap. [0003] Subsequently, CNN has achieved great success in the field of image recognition. As a high-level semantic representation, gl...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F17/30G06N3/08
CPCG06F16/583G06N3/08
Inventor 徐杰卞颖盛纾纬唐淳田野
Owner 成都星亿年智慧科技有限公司