Credential layout analysis method and device
A format and certificate technology, applied in the field of certificate format analysis, can solve the problems of cumbersome development process, uncertain results, and large workload, and achieve the effect of avoiding repeated development, reducing workload, and making small changes.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0044] see figure 1 , the present invention provides a flow chart of a method for certificate format analysis, including:
[0045] Step S1, obtaining the certificate image;
[0046] Specifically, the certificate image can be taken by a terminal device connected to a camera or a terminal device with a built-in camera, or it can be an image intercepted by analyzing a video stream or a directly stored certificate image; the terminal device can be, for example, a mobile phone or a tablet computer , PDA (Personal Digital Assistant, personal digital assistant, referred to as: PDA), etc.
[0047] Step S2, extracting the format features in the document image;
[0048] Specifically, extracting the typography feature is composed of the character gradient direction histogram feature, the inter-line distribution feature and the intra-line character inter-character feature.
[0049] Step S3, using a document recognition model to identify each of the format features, and obtain the corre...
Embodiment 2
[0055] Such as figure 2 As shown, the training flowchart of the certificate recognition model in the method for certificate format analysis provided by the present invention includes:
[0056] Step S101, collecting certificate images of different formats among similar certificates;
[0057]Among them, if the document to be analyzed is an ID card, then the document images of different versions of the ID card need to be collected; if the document to be analyzed is a passport, then the document images of different versions of the passport need to be collected; if the document to be analyzed is Bank bills, then different versions of bank bill images need to be collected; according to the different types of documents to be analyzed, different versions of the document images are selected.
[0058] Step S102, extracting all the layouts in each document image and the layout features corresponding to each layout,
[0059] Wherein, each certificate image contains multiple text lines,...
Embodiment 3
[0066] Such as image 3 As shown, the flow chart of step S2 in the method for document format analysis provided by the present invention includes:
[0067] Step S201, performing binary segmentation on the certificate image to obtain corresponding text lines;
[0068] Among them, the purpose of adopting the principle of binarization segmentation is to process the key points in the document image, remove the background by the way when segmenting the image, leave the target object of interest, and facilitate the extraction of text lines; the method of binarization segmentation specifically includes the following Three types, threshold based on pixel value, threshold based on region property or threshold based on coordinate position.
[0069] Step S202, sequentially selecting different character lines to combine to generate multiple layouts, wherein each combination is a layout;
[0070] Wherein, each text line obtained by segmentation is combined to generate a plurality of layo...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com