News classification
Contact us
- Add: No. 9, North Fourth Ring Road, Haidian District, Beijing. It mainly includes face recognition, living detection, ID card recognition, bank card recognition, business card recognition, license plate recognition, OCR recognition, and intelligent recognition technology.
- Tel: 13146317170 廖经理
- Fax:
- Email: 398017534@qq.com
Card scanning recognition technology in WeChat
It is believed that many people are interested in the function of the scanning identification card (bank card identification) of alipay, and I would like to know the principles.
In fact, this only USES the mobile terminal OCR recognition technology. After scanning the image of the phone camera, the image processing is processed by the mobile terminal OCR recognition software to extract the card number process.
But there are certainly more subtle developers who want to know. The following is a detailed description of the technical principle of bank card scanning identification. The main steps of bank card identification technology are as follows:
1. Take the picture of the color bank card: at present, video preview mode is mainly used to draw a needle for video, and a clear color image is taken. Is the phone's camera to scan interface in real time graph, the main is judged by detecting frame of bank CARDS are on the specified box, so can guarantee to get better image quality.
2. Image pretreatment: this is mainly denoising (including grayscale processing) and binary, and finally only black and white.
3. Positioning of card number: after the completion of binarization, it is necessary to locate the industry of bank card number.
4. The single character of card number: the single digit of bank card number is divided (the segmentation is also a big difficulty in card scanning identification technology).
5. Identification of card Numbers: this step is only to compare the single digit with the numerical template in the font library to the highest degree of confidence.
6. Check before output: before the credit card identification, the verification is required, mainly using the Luhn algorithm to verify the correct output. (the Luhn algorithm is described in detail here: all the bank CARDS issued by the bank are verified. The card number is a check code. It mainly conforms to the following rules:
1) add the odd bits (1, 3, 5, etc.) from the last digit of the card number.
2) start with the last digit of the card, reverse the even number, multiply it by 2 (if the product is in double digits, then subtract 9 from it) and sum it up.
3) combined the odd bit and the even number, the result should be divisible by 10. )
7: output: output bank card number string.
The whole card identification process ends. Though it feeling is simple, but there are a lot of detail processing is still need strong technical accumulation, if simply make a SDK, recognition effect is bad, it requires professional OCR techniques can you get it ready. That's why so many businesses use our bank card identification SDK.