Similar image retrieval method, device, computer equipment and storage medium
A technology of similar images and images, applied in the computer field, can solve problems such as poor applicability and achieve the effect of excellent applicability
Active Publication Date: 2022-07-01
浙江中设天合科技有限公司
View PDF5 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0005] The purpose of the embodiments of the present invention is to provide a similar image retrieval method, which aims to solve the problem that the hash function used in the existing similar image retrieval process needs to label the image in advance before learning. Technical issues with poor applicability in scenarios
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0025] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention.
[0026] In order to solve the solution in the prior art that images need to be labeled in advance in order to learn and train to obtain the hash function used in the retrieval process of similar images, the present invention proposes a deep hash network and clustering algorithm mixed training to obtain The method of unsupervised deep hashing model, and the unsupervised deep hashing model obtained by this training method realizes the similarity retrieval of images, wherein, after the image is extracted to the feature vector of the image through the feature extraction layer of the deep hashing...
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
Login to View More Abstract
The invention is applicable to the field of computer technology, and provides a similar image retrieval method, device, computer equipment and storage medium. The method includes: acquiring an image to be retrieved; processing to determine the hash code of the image to be retrieved; the unsupervised deep hashing model is generated in advance based on the clustering algorithm and iterative optimization training of the deep hashing algorithm; the deep hashing algorithm uses the pseudo-label determined by the clustering algorithm as the Optimization objective: determine a similar hash code that satisfies a preset similarity relationship with the hash code according to the hash code of the image to be retrieved, and determine the corresponding image. In the method provided by the present invention, when faced with a large number of images in the training process, the unsupervised deep hashing model does not need to label the images one by one in advance, but directly uses the clustering algorithm to perform pseudo-label annotation on the images, which has better applicability .
Description
technical field [0001] The invention belongs to the field of computer technology, and in particular relates to a similar image retrieval method, device, computer equipment and storage medium. Background technique [0002] The retrieval process of hash images generally includes two steps. The first step is to perform feature extraction on the training data set to obtain high-dimensional feature vectors of the image, and then learn these high-dimensional feature vectors through the designed algorithm, and then get the learning method. Good hash function. The second step is to extract the image in the database through feature extraction and mapping of the hash function to obtain the hash code and store it in the database. When an image is used for retrieval, the corresponding feature extraction is also performed, and then the hash function is performed to obtain the hash code of the corresponding query image, and the similarity between the hash code and the hash code in the da...
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
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/583G06F16/55G06V10/762G06V10/75G06V10/82G06N3/04G06N3/08
CPCG06F16/583G06F16/55G06N3/088G06N3/045G06F18/22G06F18/23213
Inventor 陈墨郭唐仪练智超张德龙
Owner 浙江中设天合科技有限公司