An Information Analysis Method and Model Based on Neural Network
An information analysis and neural network technology, applied in the field of information analysis methods and models based on neural networks, can solve problems such as image information analysis and detection in images that cannot be detected, improve detection angles and methods, improve detection accuracy, and have a wide range of applications. Effect
Active Publication Date: 2021-11-19
杭州鹏泰电力设计咨询有限公司
View PDF5 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0005] Aiming at the problem that the existing technology cannot analyze and detect the image information in the pictures, the present invention provides a neural network-based information analysis method and model, which uses deep learning to learn and process a large number of pictures, and can effectively extract pictures The information in the image, and then use text to compare or describe it, so that we can know in time whether the image contains harmful factors
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
[0019] An information analysis method and model based on a neural network, comprising the following steps: S01: dividing a picture and extracting candidate areas; the specific process of step S01 includes: dividing the picture by using an RPN network, and dividing the part to be extracted from the picture Mark the information box, and return the score, and compare and judge the returned score value with the preset threshold value, and the target candidate area that is greater than or equal to the threshold value is the target candidate area that meets the requirements.
[0020] S02: Use the convolutional neural network to identify and classify the extracted picture candidate areas; the specific process of step S02 includes: adding weights and weight bias coefficients before each node of the convolutional neural network, and inputting the candidate area information input by the layer After passing the corresponding weight and weight bias coefficient, it is transmitted to the hid...
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 an information analysis method and model based on a neural network, comprising the following steps: S01: dividing a picture and extracting candidate areas; S02: using a convolutional neural network to identify and classify the extracted picture candidate areas; S03: Use the cyclic neural network to match the text with the classified candidate area information; S04: Generate a text description of the picture. The substantive effects of the present invention include: wide application range, perfect detection angle and method, capable of outputting intuitive text to express picture information, speeding up the acquisition of picture information, and improving detection accuracy.
Description
technical field [0001] The invention relates to the technical field of computer vision, in particular to an information analysis method and model based on a neural network. Background technique [0002] With the rapid development of the Internet, the information that people receive is also increasing exponentially, and the ways and types of information receiving are also showing a trend of diversification. The most direct one is picture information, but there are certain disadvantages. There are all kinds of pictures on the Internet. Information, bad information in it can negatively affect people, and the spread of these things will affect us. Therefore, it is particularly important to detect the information content of these images. [0003] The invention of authorized announcement number CN103020651B discloses a method for detecting sensitive information of microblog pictures, including: establishing a sensitive word library, a font library and a color library, receiving N...
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): G06K9/32G06K9/62G06F16/51G06F16/58
CPCG06F16/5866G06F16/51G06V10/25G06F18/22G06F18/24G06F18/214
Inventor 王越胜丁靓靓
Owner 杭州鹏泰电力设计咨询有限公司