Go to MPAI-MMC V2.5 AI Modules
Function
Ref. Model
I/O Data
SubAIMs
JSON MData
Profiles
Ref. Software
Conformance
Performance
1 Functions
The Optical Character Recognition (MMC-OCR) AIM receives the image of a page as a Basic Visual Object and, optionally, a Point on it, and produces a Text Object: the lines of text found on a band of the page about the Point (at the top of the page when there is no Point), in the full width of the page and in reading order, each a Basic Text Object with its box on the page and its confidence. The box of a line is the Space-Time of the line in the Text Object: SpatialAttitude1 is its top-left corner and SpatialAttitude2 its bottom-right corner, in the pixels of the image (X to the right, Y down). The confidence of a line is in its Data Exchange Metadata.
| Receives | Page Image | The image of the page. |
| Page Point | A Point on the page. | |
| Produces | Page Text | The lines of text found, each with its box and its confidence. |
2 Reference Model
Figure 1 depicts the Reference Model of the Optical Character Recognition (MMC-OCR) AIM.
Figure 1 – Reference Model of the Optical Character Recognition (MMC-OCR) AIM
The models of the Optical Character Recognition (MMC‑OCR) AIM are data of Access. The Metadata of the AIM declares them in AccessSources (ocr-det, ocr-cls, ocr-rec, ocr-keys) by the URI each is found at – a file, or a file of an archive, named by Member – and the SHA-256 of the file. The runtime that executes the models is part of the Implementation of the AIM.
3 I/O Data
Table 1 specifies the Input and Output Data of the Optical Character Recognition (MMC-OCR) AIM. Links are to the Data Type specifications.
| Input | Data Type | Description |
|---|---|---|
| Page Image | Basic Visual Object | The image of the page. |
| Page Point | Point (optional) | A Point on the page. |
| Output | Data Type | Description |
| Page Text | Text Object | The lines of text found, each with its box and its confidence. |
4 SubAIMs
No SubAIMs.
5 JSON Metadata
https://schemas.mpai.community/MMC/V2.5/AIMs/OpticalCharacterRecognition.json
6 Profiles
No Profiles.
7 Reference Software
Not part of this specification.
8 Conformance Testing
Table 2 provides the Conformance Testing Method for the MMC-OCR AIM.
If a schema contains references to other schemas, conformance of data for the primary schema implies that any data referencing a secondary schema shall also validate against the relevant schema, if present, and conform with the Qualifier, if present.
| Receives | Page Image (Basic Visual Object) | Shall validate against the Basic Visual Object schema. |
| Page Point (Point) | Shall validate against the Point schema. | |
| Produces | Page Text (Text Object) | Shall validate against the Text Object schema. |
9 Performance Assessment
Not part of this specification.
