Smart phone name card extraction system and realization method thereof

Abstract

The invention discloses a smart phone name card extraction system and a realization method thereof. The smart phone name card extraction system comprises an image induction and enhancement module, an image detection and denoising module, a straight line detection and revision module and an image display module, wherein the image induction and enhancement module is used for converting image-text information of a name card into an image and automatically adjusting contrast ratio of the image; the image detection and denoising module is used for detecting the edges of the image with the adjusted contrast ratio through a filter and removing noises produced in the edge detection process by the filter; the straight line detection and revision module is used for detecting straight lines of the denoised image, removing the straight lines which are less than a certain threshold value of the denoised image, revising the straight lines in the image which are subjected to straight line detection and removing the straight lines which do not intersect with each other and which intersect in the angle not equal to 90 DEG; and the image display module is used for displaying image processing effects and operation interfaces. With adoption of the smart phone name card extraction system, a user using the smart phone can extract the name card accurately under the condition that the name card has small difference with a background of the smart phone or illumination is not uniform.

Description

technical field [0001] The invention relates to the field of business card extraction and recognition for smart phones, in particular to a business card extraction system for smart phones and an implementation method thereof. Background technique [0002] In business activities, people often need to exchange business cards, but it is very inconvenient to carry business cards with them. Therefore, people use the current smart phone function to store the business card in the smart phone in the form of a picture, so that the user only needs to query the information stored on the mobile phone when querying the information on the business card. [0003] In the prior art, there are many methods for extracting target business cards, the main ones are threshold method, edge extraction algorithm, SUSAN detection operator, watershed segmentation algorithm and so on. Among them, thresholding methods can generally be divided into two categories: global thresholding and local thresholdi...

AI Technical Summary

Problems solved by technology

[0008] In the image segmentation algorithm, the threshold method is an effective and simple image segmentation algorithm, but for image processing under complex background conditions, the threshold method is difficult to accurately extract the target image, especially when the image is unevenly illuminated , this method will fail completely, which will seriously affect the extraction of business cards. In addition, when the background and foreground brightness values ​​are similar, the threshold value is difficult to determine accurately, and this method will also fail at this time

In the existing edge detection algorithms, such as Roberts operator, Sobel operator, Prewitt operator, Laplacian operator, etc., there is a common problem, that is, the image is required to be evenly illuminated, and when the image is not uniformly illuminated, it will cause some problems. False targets appear. In the actual process, since the images captured by smartphones are taken at close range, generally less than 5 inches, the shadows cast by the mobile phone and hands will cause uneven illumination of the image.

In addition, when the background and the foreground are not much different, because the threshold is difficult to select, the method will also fail at this time

Images