Text Means Remove Text

Introduction

The following issues remain.

Schema:

------The previous teleconference ended here------

GP6:

ScanDocument.ScanDocumentTicket.ScanDocumentProcessing.(ImageDocumentProcessing).NumberUp

ScanDocument.ScanDocumentTicket.ScanDocumentProcessing.(ImageDocumentProcessing).PresentationDirectionNumberUp

Since we are not printing this must refer to the number of scan areas on a single page. But I thought it was agreed that there would only be on scan (area) per page.

<PZ>Advanced scanners can perform this transformation</PZ>

GP7:

ScanDocument.ScanDocumentTicket.ScanDocumentProcessing.(ImageDocumentProcessing).PageRanges

Does this have meaning in scanning? Does the user put on a stack of page and then states scan the 12, 24 and 27 page. I believe the user will only put on the pages they want to scan. Or is this simple the NumberScanPages; that is the number pages scanned. (I noticed you use OutBoundCount which I think is NumberScanPages.)

<PZ>PageRanges moved to appropriate services</PZ>

BW7:

I do not follow why page ranges does not make sense for scan. It seems a useful attribute to me.

GP8

ScanJob.ScanJobStatus.CompletedFinisherJobs

I don’t understand what this is. I don’t associate a “finisher” with scanning only printing.

<PZ>Do we need to subclass Monitoring?</PZ>

GP9:

ScanJob.ScanJobStatus.ScanServiceCounters.FullColorImages

Why is the word “Full” added. The only other type image is MonoChromeImages. I suggest just ColorImages (versus FullMonoChromeImages). Otherwise, you need to extent the list to something like

8BitColorImages …. 24BitColorImages …. 48BitColorImages … IndexColorImages ?????

<PZ>Inherited from Print semantics to differentiate between full color from highlight color</PZ>

GP10:

Since I do not see a variable for Rotation, I assume that Orientation is recording the user desire for rotation of the output scanned image. But scanner can do orientation (rotation) in two ways. All scans for a specific scanner will physically be in a single orientation but the Scan Service or Scan Driver can perform orientation (rotation) by either setting an orientation (rotation) attribute in the image file and the reading application will perform the actual orientation (rotation) change. The other way is the Scan Driver performs the actual orientation (rotation) the image content. Does there need to be an OrientationType.

<PZ>Rotation (0,90,180,270) added to DocumentProcessing</PZ>

GP11:

The concept of Magnification (as a capability and as an option) is not typical used in scanning as it is for printing. In scanning, the concept of magnification is replaced with changing the resolution (dpi) of the scan. If you want a bigger scanned image then increase the resolution. And you don’t really change the resolution to get a bigger picture but to get a better quality image. Yes, cameras have a magnification option but it is either physical-optical or software-optical magnification. “Copy” has a magnification option but that is software magnification of whatever the scan image “size” (resolution) is for the printer.

GP12:

Is the capability “OutputDocumentSize” correct or it is supposed to be “InputDocumentSize”; this is related to the AFD and Platen width/height dimensions. The would also mean there would have to be two InputDocumentSize’s since the platen and AFD may have different size parameters.

GP13:

I believe the capability “Sharpness” refers to a capability of high-end scan for “Focusing”.Typical scaners do not have Sharpness nor Focusing; this is done by a post-processing application. I would suggest changing the “Sharpness” to “Focus” or “Focusing” to more accurately map to the physical operation.

GP14:

I did not see a capability for the “light source” this can be changed in scanning and can be done for the document type (paper, transparencies, negatives).

GP15:

I did not see a capability for “document type”. Today, even low-end scanners can scan paper, transparencies, photo-negatives, photos, etc. Each of these may affect other scanning attributes. Example, scanning photo-negatives are done on a scan area template or a single photo-negative in the center of the platen. And, yes, office people scan photo-negative, photos, transparencies.

<PZ>It is OriginalType with values of Auto, Text, Photo, Magazine, Halftone, Mixed</PZ>

GP16:

Schema & Specification

[optional] AutoSkewCorrection [Boolean: On, Off]

AutoSkewCorrection will provide detection and correction of small skew orientation error from the media loading from the MFD or placement on the platen by the user.

<PZ>Added to schema</PZ>

GP17:

Schema & Specification

[optional]DocumentPageElementDetect[Boolean:On, Off][Type: Image, Graphs, Text, Barcode, Form-Elements (CheckBox, etc), All]

The ScanService may provide the function of segmenting the individual scanned document pages into document element components; such as, images, graphs, text, barcodes, etc.. Using the multiple scan regions per scanned document page, ScanService can store elements in separate files.

GP18:

Schema & Specification

Another possibility since image processing is included is recognition (OCR, barcodes, forms-elements)

[optional]DocuemntPageRecognition[int: REC_OCR, REC_BARCODE, REC_FORMS (or something like this)]

The ScanService may provide the function of segmenting the individual scanned document pages into document element components and provide recognition of the element components; such as text, barcodes, form-elements.

PZ1:

Discuss Configuration, Components or Subunits (My preference is Configuration)

PZ2:

We need to discus and resolve the unification of coordinate system across the services and subunits. What should the element name be throughout the schema?

GP20:

In the current ScanService specification we have image processing functions such as brightness, contrast, sharpness, etc. My assumption is that network scanning is akin to batch processing versus a direct connected MFD (/scanner) which would provide a more interactive mode of operation. Therefore, I wonder how an end user will set the values for brightness, contrast, sharpness, etc in a non-interactive mode. Any of these image processing settings may change from page-to-page in a single document and from document-to-document. For a non-interactive (batch) scan processes, I believe that the image processing function should only be “Auto”; i.e. Auto-Brightness, Auto-Contrast, Auto-Sharpness, etc.At a minimum, there has to be an “Auto” setting.This way optimization is done on a page-to-page/document-to-document basis without user interaction. It may be desired to report (return) the “Auto” setting that was actually used for each function on a page-to-page basis

<PZ>In the current model there is a choice between autoexposure and setting brightness contrast and sharpness. Is this sufficient or should auto be at the individual level? What are the semantics if one of the elements is omitted?>

GP21:

My background in physics, optics and image processing has always used the term monochrome to mean “an image is presented as different shades of gray from black to white”; with the more general definition to be “an image is presented as different shades of a single color from the color to white”. After searching the web (do a Google search using the key words “define: monochrome”) you will discover that you can find just about any definition you want; including:

Monochrome Printer - A monochrome printer can only produce an image consisting of one color,

usually black. A monochrome printer may also be able to produce graduations

of tone of that color, such as a grey-scale.

So this mean monochrome is both B/W and grey (gray)!!!!!!!!!!!!!!

I believe you will find that the term monochrome migrated to it current association as a binary B/W representation from how people have talked about photos. Photos have always been denoted as either being “Black/White” or “Color”. This is done to distinguish two big classes of photos. However, the B/W photos (images) are actually gray tone (black monochrome) not binary B/W.

People also to refer to monochrome displays; but these are not binary B/W either, since the display can display a range of gray (green, red, brown) values but adjusting the intensity of the electron beam.

I would like to propose the following.

Binary Image – An image composed of individual pixels having only two possible states and represented by a single binary value. Example: Black and White, Red and White, Blue and White, Blue and Red.

Monochrome Image – An image composed of individual pixels having n number of possible states that corresponds to a graduation of tone of a single color and represented by 1 or more bytes per pixel. Examples: 8-Bit-Gray-Tone (256 states going from black to white) and 16-Bit-Sepia (65536 states going from brown to white)

Color Image – An image composed of individual pixels having n number of possible states representing Hue, Value and Intensity where the Hue, Value and Intensity are represented as either indexed individual color values or color space coordinates. Examples: 8-Bit-RGB (256 possible indexed states), 24-Bit-RGB ((8,8,8) = 16 million possible states), 48-Bit-RGB ((16,16,16) = 2.8 E14 possible states), CMYK, HSV

BW comment:

I apparently missed a discussion of the meaning of monochrome, but I consider the term to include gray-scale. After all, it refers to a single color (χρωμα) but does not refer to shade/density/intensity. I consider a monochrome a characteristic as distinguished from polychrome, or full color or spot color.

However, the definitions, need some discussion. This is especially true of digitally represented color images.

Further, I suggest that we specify that the definitions refer to the conventional digital representations of images; there are also hardcopy images, the characteristics definition of which is beyond the scope of the MFD semantics activity.

NS1:

Mistake in MediaSizeNameExtensionPattern

<xsd:simpleType name="MediaSizeNameExtensionPattern">

<xsd:restriction base="xsd:NMTOKEN">

<xsd:maxLength value="255"/>

<xsd:pattern value="(custom_max|custom_min|custom|na|asme|roc|oe)_[\p{Ll}\d][\p{Ll}\d\-]*_\d+(\.\d+)*x\d+(\.\d)*in"/>

<xsd:pattern value="(custom_max|custom_min|custom|iso|jis|jpn|prc|om)_[\p{Ll}\d][\p{Ll}\d\-]*_\d+(\.\d+)*x\d+(\.\d)*mm"/>

</xsd:restriction>

</xsd:simpleType>

Please watch the last characters of both patterns.

With the current schema one would have to enter values like “custom_small_3.21x4.3.2in” instead of “custom_small_3.21x4.32in” (watch the dots).

So the part about the width is correct, the part about the height is wrong.

BW1:

Default scan ticket: since a scan template is defined as an unbound scan ticket, is not the term default scan template more consistent?

BW2:

(Note: from meeting minutes implication of indenting) Are “state, State Reason” etc considered part of Scan ID? Also It would seem that there should be common element in all the scan service headings so that they can all be applied to the same scan service. That is, should not some sort of Scan Service ID should be common.

<PZ>There are common service status elements that include the above elements (See updated schema)</PZ>

BW3:

“Service Type” is eliminated because I is said to be in Scan Service Status. Neither do I see service type in Scan Service nor would I consider type to be a status. Of course, it is not clear what the values of Service type are. If they are scan, print, fax etc, then the type may be regarded as inherent in the group names (unless the same group names are used for all services.)

<PZ>Each Service is defined in its own namespace at this time (See Server.Services in Schema).</PZ>

BW4:

Scan Service Capabilities – Would it not be clear if these various elements were followed by “supported” ? e.g., Number-Up Supported

BW5:

Since the term “Must Honor” is used exclusively in the discussion, would not “Must honored element list” be more appropriate than JobMandatorELements?

<PZ>The ‘MustHonor’ attribute comes from existing Web Service and XML schema practices. The “JobMandatoryElements” is right out of IPP/PWG SM v1</PZ>

BW6:

For my own edification, are there clear definitions of element and element attribute? The previous paragraph talked about must honored elements and this paragraph talked about must honored attributes.

<PZ>”MustHonor” is the only defined attribute in the schema. Everything else is represented as an element.</PZ>

BW8:

We need some expertise do define the scanner subunit. For example physical constituents of the CIS vs CCD based scanner are quite different.