Indonesia Card Anti-fake
Use the deep learning algorithm to detect whether the face in the certificate is PS, whether the certificate is a printed certificate, replayAttackd certificate, and the blur of the certificate,the illumination of the certificate
Code example:
curl --location --request POST 'https://api.ekycpro.com/v1/id_card/antifake' \--header 'Content-Type: application/x-www-form-urlencoded' \--header 'X-API-Key: AbcdEfgh' \--data-urlencode 'img=BASE64_ENCODE_VALUE'
Response for the input face is ps
{ "status": "OK", "message": { "blur": { "isBlur": "no", "score": 0 }, "ps": { "isPs": "yes", "score": 95 }, "print": { "isPrint": "no", "score": 1 }, "replayAttack": { "isReplayAttack": "no", "score": 10 } }}
Response for the input card face is not ps
{ "status": "OK", "message": { "blur": { "isBlur": "no", "score": 0 }, "ps": { "isPs": "no", "score": 8 }, "print": { "isPrint": "no", "score": 1 }, "replayAttack": { "isReplayAttack": "no", "score": 10 } }}
Response for the input card is blur image
{ "status": "OK", "message": { "blur": { "isBlur": "yes", "score": 61 }, "ps": { "isPs": "no", "score": 8 }, "print": { "isPrint": "no", "score": 1 }, "replayAttack": { "isReplayAttack": "no", "score": 10 } }}
Response for the input card is print image in black and white
{ "status": "OK", "message": { "blur": { "isBlur": "no", "score": 1 }, "ps": { "isPs": "no", "score": 8 }, "print": { "isPrint": "yes", "score": 100 }, "replayAttack": { "isReplayAttack": "no", "score": 10 } }}
Response for the input card is replayAttackd image
{ "status": "OK", "message": { "blur": { "isBlur": "no", "score": 1 }, "ps": { "isPs": "no", "score": 8 }, "print": { "isPrint": "no", "score": 0 }, "replayAttack": { "isReplayAttack": "no", "score": 10 } }}
Response for the input card is too dark or too bright
{ "status": "OK", "message": { "blur": { "isBlur": "no", "score": 1 }, "ps": { "isPs": "no", "score": 8 }, "print": { "isPrint": "no", "score": 0 }, "replayAttack": { "isReplayAttack": "no", "score": 10 } }}
Response for recognition fail
{ "status": "FAIL", "message": "check input image and retry"}
Response for invalid request parameters
{ "status": "INVALID_REQUEST", "message": "check request params"}
Response for invalid image format
{ "status": "IMAGE_INVALID_FORMAT", "message": "image format not support"}
Response for image size larger than 2M
{ "status": "IMAGE_INVALID_SIZE", "message": "image larger than 2M"}
Response for server error
{ "status": "INNER_ERROR", "message": "Inner error, please retry later"}
HTTP request
POST https://api.ekycpro.com/v1/id_card/antifake
Request parameters
parameter | description |
---|---|
img | string , base64 encoded image. It is recommended that the image be less than 200KB, so it will be returned within 2 seconds, otherwise the return time will be longer. |
Response format
The scores below are reference values. It is recommended to adjust the scores in actual use.
fields | description |
---|---|
isPs | yes face is ps, no face is not ps |
ps score | this is the confidence level of the ps face, the range is [0, 100]. The default is greater than 80 points is ps face |
isBlur | yes is blur, no is not blur |
blur score | this is the confidence level of the blur card, the range is [0, 100]. The default is greater than 30 points is blur card |
isPrint | yes is print image in black and white, no is not print image |
print score | this is the confidence level of the print image, the range is [0, 100]. The default is greater than 70 points is print image |
isReplayAttack | yes is a replayAttackd image, no is not replayAttackd image |
replayAttack score | this is the confidence level of the replayAttack image,the range is [0, 100]. The default is greater than 60 points is replayAttackd image |
Status code
status | description |
---|---|
OK | charge , success |
FAIL | charge , image recognition error, please check input image |
INVALID_REQUEST | free , invalid request parameters |
IMAGE_INVALID_FORMAT | free , invalid image format, image format should be one of jpeg/jpg/png/bmp |
IMAGE_INVALID_SIZE | free , invalid image size, should be less than 2M |
INNER_ERROR | free , server error |