Photocopying – Projection printing and copying cameras – Illumination systems or details
Reexamination Certificate
1992-08-26
2001-06-12
Metjahic, Safet (Department: 2851)
Photocopying
Projection printing and copying cameras
Illumination systems or details
Reexamination Certificate
active
06246464
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to illumination systems and more particularly to associated techniques yielding intense, Lambertian beams.
BACKGROUND, FEATURES
Document processing is a well developed art today; for instance, witness the processing of bank checks, remittance documents and other documents of finance handled by the millions every day in banks all over the world. To handle this incredibly large volume of paper, banks rely on high speed document processors that feed thousands of documents per minute past various processing stations. [E.g. Unisys Corp. has DP1000, DP1800 and 9195 high speed document processors that handle 1000, 1800, and 2600 documents per minute, respectively.] Processing a bank document typically involves some or all of the following steps: e.g., document-feed, reading magnetic ink recognition characters (MICR), reading optical recognition characters (OCR), printing endorsements, microfilming, sorting/stacking (routing documents to pockets), and imaging (see FIG.
1
A).
The imaging task typically involves: Acquisition, processing, compression, storage, transmission, display, printing, and archive of images. This invention involves producing an image platform apt for integration with a high speed sorter for acquisition, processing, and compression; with storage retrieval modules for storage and image management; with image workstations for display; image printers for print; with optical disk subsystems for archive, and with means to transmit images, point to point, and local area networks (see FIG.
1
B). Mainframe computer “hosts” can link multiple imaging document processors, and associated storage retrieval modules and other peripherals to create a large scale system (e.g. see FIG.
1
C).
For image-acquisition, we prefer to use a camera which provides the system with images of the front and rear of the processed documents; e.g. facsimile images which are adequate for a customer's needs.
THE PAPER MOVER
To fix the parameters of camera design, one must consider the mechanical transport that moves documents past the camera sensor. The document transports here contemplated can propel paper past processing stations at track speeds of 100 to 400 inches per second (ips). The 5 mil thick documents (0.005″) move along a track that is made 70-90 mils wide (TRACK GAP) in order to minimize jams and accommodate staples and other foreign objects that may be still attached to a document.
Document size determines the field of view required to image a maximum-height document. Most documents here contemplated (e.g. checks) are 2.75 to 4.75 inches high and between 6 and 10 inches long. A document may exhibit “skew” as it travels through the transport and so require a field of view higher and wider than the actual document size.
Camera design is also affected by a transport's degree of “document control”, i.e. how much excess paper-motion (within track-width) is permitted in the area of image acquisition. Control techniques range from costly vacuum belt systems that minimize paper movement across the track to 20 mils or less, to less restrictive approaches that allow more document motion, but cost less. The tradeoff between minimum vs. maximum document control involves cost, complexity and—most important to the camera designer—the depth of illumination, and the amount of light required to properly illuminate the document in the image acquisition area.
In the instant embodiments we opt for minimal document control to reduce cost of the lower overall system and to increase transport reliability (minimize jams); thus, we must maximize the output of the illumination source—as a salient object hereof.
ILLUMINATION SOURCE
Selection of the illumination source is influenced by the depth of illumination desired and by the image sensor one selects (this dictates the sensor light input required). As documents are propelled past the image sensor at from 100 to 400 ips, there is little time to accumulate enough photons to form an acceptable image (i.e. to meet signal to noise and dynamic range requirements) so one tends to maximize light input.
There are some elegant sensor arrangements, such as time delay integration (TDI) devices, that reduce input light requirements, but these bring along their own unique set of needs such as synchronization of document motion. Also they tend to be “custom” arrangements using components with limited commercial availability and to involve package configurations (electrical and mechanical) that are non-optimal for our contemplated applications. Our image sensor of choice is a “sensor chip”, such as the Reticon RL1288D, or a like chip which provides compatible output data rates of up to 80 MB/s and is priced attractively (e.g. the cost of a custom chip).
The type of light reaching our image sensor is very important for proper acquisition of color information from a document. We expect to satisfy the “eyeball test”: i.e. “If I can see it on the document, I expect to see it in the image.” Although “downstream processing” (after the camera acquires the image) can remove information, it cannot recover information that the camera doesn't acquire. Therefore it is important that the camera provide a representation, or facsimile, of the original document that is as faithful as possible.
Thus, the color response (spectral response) we require of our image camera will be the response of the human eye, i.e. “photopic response”. Taking “photopic response” as our norm, we must factor-in the frequency response of our sensor, the output spectrum of our light source spectrum, and the effects of our optical-path-elements. To account for all these we interpose a photopic filter in the light path to the sensor—i.e. to emulate the desired photopic response.
Although photopic response is our “baseline”, it is also useful to further shape sensor (spectral) response. For example, we find, we have found that a slight “red shift” (moving sensor response curve toward the red), with modifications to the response curve edge rates, gives optimal results for applications requiring “drop-out” of red inks.
GENERAL CONSIDERATIONS
One will optically-couple his light source and image sensor with suitable optics (e.g. lenses, mirrors), and prefer to package all this as an integral illuminator-camera unit which can be integrated into a pre-existing document processor. [E.g. a front and rear camera system]. The light from the illumination source will normally travel through a series of mirrors, lenses and fiber optics prior to reaching the document. A preferred lamp is a high output, tungsten-halogen filament type, less preferred is a xenon arc lamp that is highly efficient in the visible spectrum; and, for “slower” document speeds, certain fluorescent lamps are acceptable. All the optics are optimized with reflective (AR) coatings to minimize light loss.
With a passing document so illuminated, front and rear, it can present reflected-light images to the image sensor via relay optics and an “imaging lens” specially selected to produce picture elements (or “pixels”) that are properly sized for downstream image processing algorithms. An important characteristic of such imaging lenses is their magnification tolerance; i.e. the sensitivity of subsequent processing to pixel-size variation. We have taken a conservative approach here to insure highly repeatable pixel dimensionality through appropriate selection of the image lens focal length. Depth of focus or the ability to stay in focus throughout the track gap is another key design parameter.
LIGHT-TO-VOLTAGE TRANSLATION
Having brought a reflected image (e.g. from front and rear of document) to the image sensor, we next need to convert this into a video signal that can be understood by our imaging electronics. The electrical signals output by our sensor chips need to be conditioned, amplified, filtered and then converted to a digital form by an analog to digital (A/D) converter.
Here, the linearity of the signal is a consideration. [Note: the response of camera electronics to
Bakker Johan P.
Catchpole Clive E.
Concannon David J.
Copenhaver Gary B.
Rourke Robert T.
Cass Nathan
Dalakis Michael
Metjahic Safet
Starr Mark T.
Unisys Corporation
LandOfFree
Illumination system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Illumination system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Illumination system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2513229