Television station logo identification system based on deep learning

A deep learning and station logo identification technology, applied in the field of computer vision, can solve the problems of overlapping station logos, difficult similar colors and transparent station logos, low recognition rate, etc., to achieve high real-time effect.

Inactive Publication Date: 2017-04-26
SICHUAN CHANGHONG ELECTRIC CO LTD
View PDF2 Cites 17 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] Most of the existing station logo recognition methods have a low recognition rate for the following situations or do not consider the following situations at all: (1) similar colors and transparent station logos are not easy;

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
  • Television station logo identification system based on deep learning
  • Television station logo identification system based on deep learning
  • Television station logo identification system based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0056] Take training a model containing 80 logos as an example to illustrate the training process of the logo recognition model.

[0057] (1) Collect samples of each type of logo, index: the number of samples is 3000 for each type, and the collection time interval between each sample is 3s;

[0058] (2) Delete samples that are not suitable for training in the sample;

[0059] (3) Select a sample that is close to a black background in each type of station logo, and segment the station logo foreground image and the station logo itself image;

[0060] (4) Artificially synthesized samples, 2000 samples per category were synthesized according to the fixed station logo position method and the arbitrary station logo position method;

[0061] (5) Train the basic model on the Imagenet dataset;

[0062] (6) The actual collected Taiwan logo samples and artificially synthesized Taiwan logo samples are used as the final samples, and all samples are randomly divided into 4:1, which are re...

Embodiment 2

[0067] (1) Deploy the model and recognition results to the external Changhong 5508 core board;

[0068] (2) Connect the signal source to the input end of the core board, and connect the output end of the core board to the TV end;

[0069] (3) Set the recognition interval time to 5s;

[0070] (4) If the currently intercepted image is Sichuan International Channel, then the output result of the model is an 80-dimensional vector, and the program will calculate the position of the largest score in the vector. If the score is greater than 0.95, then the recognition result of the program is the station logo information represented by the maximum position, that is, the Sichuan International Channel; otherwise, it will not be output.

[0071] To sum up, the present invention realizes a high-efficiency and high-precision real-time TV station logo recognition method and system through local program collection of station logo samples, sample screening, artificial synthesis of samples, a...

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 relates to a computer vision technology, and discloses a television station logo identification system based on deep learning to improve the ability of station logo identification. The system comprises a sample collection module used for collecting station logo samples, a sample screening module used for screening samples to remove inappropriate samples, a station logo segmentation module used for segmenting a station logo from a background, a sample synthesis module used for artificially synthesizing different samples, a model training module used for training a station logo identification model based on the collected samples and the artificially synthesized samples, and a station logo identification module used for identifying station logos based on the identification model. The system of the invention is suitable for identifying normal station logos, rebroadcast station logos and extremely similar station logos.

Description

technical field [0001] The invention relates to computer vision technology, in particular to a TV station logo recognition system based on deep learning. Background technique [0002] With the vigorous development of broadcasting media, broadcasting and television have penetrated into every aspect of daily life and work. Considering the important role of TV station logos in distinguishing TV stations, it is of great significance to realize computer automatic recognition of TV station logos. 1. The station logo of a TV station is a distinctive sign that distinguishes many TV stations from other TV stations. It can be used to protect their commercial interests, and it is also an important manifestation of whether there is illegal insertion, suspension or black screen of the TV signal; 2. The station logo identification of a TV station can be used for cable pay channels Third, it helps the image processing tool to remove the logo, thereby improving the quality of the video. 4...

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
IPC IPC(8): G06K9/32G06K9/34G06K9/62
CPCG06V10/25G06V10/267G06V2201/09G06F18/24G06F18/214
Inventor 伍强刘明华
Owner SICHUAN CHANGHONG ELECTRIC CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap