Unlock instant, AI-driven research and patent intelligence for your innovation.
Logo detection method and system based on deep learning
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A technology of deep learning and detection methods, applied in the field of image processing, which can solve problems such as high requirements, large amount of calculation, and time-consuming
Pending Publication Date: 2021-02-05
SHANDONG NORMAL UNIV
View PDF6 Cites 2 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
The image detection work is relatively complicated, because there are various scenes in life, and the Logo data has the characteristics of many categories, similarities between categories, small objects and deformation, so Logo image detection is a big challenge
In addition, deep learning has high requirements for GPU and other hardware, and the amount of calculation is large and time-consuming.
[0006] To sum up, to balance the accuracy and speed of Logo detection, there is still a lack of effective solutions
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
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0031] Such as figure 1 As shown, this embodiment provides a Logo detection method based on deep learning, including:
[0032] S1: Build a Logo detection model based on the improved upsampling operation and loss function;
[0033] S2: Predict the upsampling kernel according to the input feature map, reorganize the target features according to the predicted upsampling kernel, and obtain the recombined features;
[0034] S3: Train the Logo detection model based on the recombined features, and use the trained Logo detection model to detect the Logo image to be tested.
[0035] In this embodiment, crawler technology is used to crawl various Logo image data on multiple websites; specifically, input keywords, such as clothing brands, food brands, electronic brands, etc., each type obtains many brands, and crawls through crawler technology. The website crawls the logo image of each brand.
[0036] In this embodiment, data cleaning and labeling are performed on the acquired Logo im...
Embodiment 2
[0065] This embodiment provides a Logo detection system based on deep learning, including:
[0066] A model building module configured to build a Logo detection model based on an improved upsampling operation and a loss function;
[0067] The upsampling module is configured to predict the upsampling kernel according to the input feature map, and reorganize the target features according to the predicted upsampling kernel to obtain the recombined features;
[0068] The detection module is configured to train a Logo detection model based on the recombined features, and use the trained Logo detection model to detect the Logo image to be tested.
[0069] It should be noted here that the above-mentioned modules correspond to steps S1 to S3 in Embodiment 1, and the examples and application scenarios implemented by the above-mentioned modules and corresponding steps are the same, but are not limited to the contents disclosed in Embodiment 1 above. It should be noted that, as a part o...
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 discloses a Logo detection method and system based on deep learning. The Logo detection method comprises steps of a Logo detection model being constructed based on improved up-sampling operation and a loss function; predicting an up-sampling kernel according to the input feature map, and recombining the target features according to the predicted up-sampling kernel to obtain recombined features; and training a Logo detection model based on the recombination features, and detecting the Logo image to be detected by using the trained Logo detection model. Based on a deep learning detection model and in combination with a new up-sampling operation and a loss function, a large receptive field can be provided during feature recombination, a recombination process is guided accordingto input features, the whole process is light in weight, good balance is achieved in speed and precision, and problems of divergence and the like in the training process are avoided; and the regression process becomes more stable.
Description
technical field [0001] The present invention relates to the technical field of image processing, in particular to a logo detection method and system based on deep learning. Background technique [0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art. [0003] With the rapid development of computer vision, target detection and recognition technology has a wide range of applications in multimedia, transportation, medical imaging and many other aspects. Due to the rapid development of Internet technology, a large amount of picture data is stored on the network. The information contained in these pictures is very meaningful. For example, the Logo contained in the picture, as a relatively important logo of the brand, plays a very important role in commercial advertisements. role. By accurately identifying the advertiser's product logo and judging whether it is an illegal product b...
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
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to View More 
Login to View More 
