FIELD: data processing; computer engineering.
SUBSTANCE: training device according to one embodiment includes a receiving unit, an extraction unit, a plurality of processors and an identification unit. Receiving unit is configured to obtain an image of a sheet of paper, which is a captured image of a sheet of paper. Extraction unit is configured to extract a plurality of images of primary features having different objects to be recognized from an image of a sheet of paper obtained by the acquisition unit. Multiple processors are configured to perform corresponding convolution and sub sample processes over multiple images of primary features extracted by extraction unit, to form multiple images of secondary features having different objects to be recognized. Identification unit is configured to sequentially update and train a set of parameters for identifying the type of sheet of said sheet of paper based on result of process of combination over each of multiple images of secondary features formed by multiple processors.
EFFECT: disclosed are a training device, a paper sheet identification device and a paper sheet identification method.
15 cl, 8 dwg
| Title | Year | Author | Number |
|---|---|---|---|
| PAPER SHEETS PROCESSING SYSTEM AND PAPER SHEETS PROCESSING METHOD | 2018 |
|
RU2692879C1 |
| DEVICE TO PROCESS PAPER SHEETS | 2012 |
|
RU2488887C1 |
| SHEET PROCESSING SYSTEM AND SHEET PROCESSING METHOD | 2016 |
|
RU2682578C2 |
| PROCESSING DEVICE OF PAPER SHEETS | 2014 |
|
RU2562791C1 |
| DEVICE FOR PROCESSING OF PAPER SHEETS | 2013 |
|
RU2555802C2 |
| DEVICE FOR PAPER SHEET PACKAGING | 2017 |
|
RU2668325C1 |
| SHEET PROCESSING SYSTEM AND METHOD FOR SHEET PROCESSING | 2016 |
|
RU2643933C2 |
| SHEET PROCESSING SYSTEM | 2015 |
|
RU2697620C2 |
| SHEET PROCESSING DEVICE AND SHEET THICKNESS DETERMININATION METHOD | 2016 |
|
RU2622478C1 |
| PAPER RECOGNITION DEVICE, PAPER PROCESSING DEVICE AND PAPER RECOGNITION METHOD | 2007 |
|
RU2436165C2 |
