Haikang has achieved the world's best score in the ICDARRobust Reading competition dataset by means of OCR (Optical Character Recognition) technology in the first article in the "Hikang Deep Learning "Text Recognition Technology International Competition". In the future, this security giant will continue to improve the performance of OCR technology, and want to promote the machine's intelligence and automation level, applied to video surveillance, intellisense and scene understanding. So what is OCR technology? What are the application scenarios?
OCR intelligent identification technology
OCR intelligent recognition technology is the OCR recognition software. The OCR software extracts and recognizes the text in the picture and converts it into searchable data. However, the OCR intelligent identification technology is actually very detailed. The OCR intelligent identification technology is introduced below.
1. Document OC identification
This type of PC is based on the beginning. In recent years, it has been developing to the mobile side. There are mainly SDKs for android and ios platforms. Currently, there are ID card identification, driving license identification, driver license identification, passport identification and so on.
2. Bank card OCR recognition
This kind of technology is a technology that has been launched since 2014. It is mainly used for mobile payment binding. It is a very technically-subdivided OCR technology. Some APPs are already in use, such as Lakara, Pleasant Loan, Alipay. WeChat and so on.
3. Business card OCR recognition
This type of technology is also very mature, and there are many apps for business card management on the market, and most of them have already used such technologies.
4. Document OCR recognition
In fact, the earliest time OCR was the main identification document, mainly for books, newspapers, etc., to electronically these paper documents. However, this type of technology is mainly developed on the basis of scanners, and the recognition rate is also very high. The OCR recognition of the document on the mobile side has also been successfully applied in recent years.
5. Bill class OCR recognition
This kind of technology is mainly for bills of different formats, mainly based on the template mechanism. It is necessary to customize different identification elements for different bills. This technology is also called element recognition OCR. The earliest use is the banking industry, mainly in the banking industry. There are very successful applications in the bank's post-supervisor.
6. Handwritten OCR recognition
In fact, handwritten OCR can be divided into two categories, one is written handwriting recognition, this kind of technology is not well applied, only because everyone's handwriting habits are not the same. At present, it is possible to recognize the numbers and letters that are handwritten, because these differences are not too big, and the handwritten man is still very difficult. The second category is online handwriting technology: online handwriting technology has been very well applied, and various handwriting input methods are based on this technology. Some people may ask, at the same time handwritten OCR intelligent recognition technology, in order to achieve very good handwriting recognition, there are actually two points, one is that online handwriting is single word recognition, no need to do the split, the most difficult thing in OCR is not the identification, and Is a character segmentation. Second, because the online handwriting technology is a single word selection, some people make judgments. So the application is very wide.
7. License plate OCR recognition
The license plate recognition technology believes that everyone will not feel strange, intelligent transportation, community parking lot, etc., have a good application, the principle of license plate recognition is actually the technology of OCR recognition of the license plate, and then the process of comparison. It is also a fairly mature technology.
OCR intelligent identification technology
OCR intelligent recognition technology is the OCR recognition software. The OCR software extracts and recognizes the text in the picture and converts it into searchable data. However, the OCR intelligent identification technology is actually very detailed. The OCR intelligent identification technology is introduced below.
1. Document OC identification
This type of PC is based on the beginning. In recent years, it has been developing to the mobile side. There are mainly SDKs for android and ios platforms. Currently, there are ID card identification, driving license identification, driver license identification, passport identification and so on.
2. Bank card OCR recognition
This kind of technology is a technology that has been launched since 2014. It is mainly used for mobile payment binding. It is a very technically-subdivided OCR technology. Some APPs are already in use, such as Lakara, Pleasant Loan, Alipay. WeChat and so on.
3. Business card OCR recognition
This type of technology is also very mature, and there are many apps for business card management on the market, and most of them have already used such technologies.
4. Document OCR recognition
In fact, the earliest time OCR was the main identification document, mainly for books, newspapers, etc., to electronically these paper documents. However, this type of technology is mainly developed on the basis of scanners, and the recognition rate is also very high. The OCR recognition of the document on the mobile side has also been successfully applied in recent years.
5. Bill class OCR recognition
This kind of technology is mainly for bills of different formats, mainly based on the template mechanism. It is necessary to customize different identification elements for different bills. This technology is also called element recognition OCR. The earliest use is the banking industry, mainly in the banking industry. There are very successful applications in the bank's post-supervisor.
6. Handwritten OCR recognition
In fact, handwritten OCR can be divided into two categories, one is written handwriting recognition, this kind of technology is not well applied, only because everyone's handwriting habits are not the same. At present, it is possible to recognize the numbers and letters that are handwritten, because these differences are not too big, and the handwritten man is still very difficult. The second category is online handwriting technology: online handwriting technology has been very well applied, and various handwriting input methods are based on this technology. Some people may ask, at the same time handwritten OCR intelligent recognition technology, in order to achieve very good handwriting recognition, there are actually two points, one is that online handwriting is single word recognition, no need to do the split, the most difficult thing in OCR is not the identification, and Is a character segmentation. Second, because the online handwriting technology is a single word selection, some people make judgments. So the application is very wide.
7. License plate OCR recognition
The license plate recognition technology believes that everyone will not feel strange, intelligent transportation, community parking lot, etc., have a good application, the principle of license plate recognition is actually the technology of OCR recognition of the license plate, and then the process of comparison. It is also a fairly mature technology.
Common Mode Coil For Automotive Product,Magntic Inductor Coil,Emi Common Mode Choke,CMC filter choke inductor
IHUA INDUSTRIES CO.,LTD. , https://www.ihua-transformer.com