Log time processing and stitching system

Details Log time processing and stitching system Log time processing and stitching system

1998-11-20

2003-07-15

Lee, Cheukfan (Department: 2622)

Facsimile and static presentation processing

Natural color facsimile

Scanning

C358S509000, C358S487000, C358S475000

Reexamination Certificate

active

06594041

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates generally to electronic film developing and more particularly, to apparatus and methods for determining the optimum exposure of pixels in a time-scanned film image.
BACKGROUND OF THE INVENTION
Chemical film development traditionally involves the chemical manipulation of an entire image which has been recorded on conventional film. Conventional film includes multiple layers of varying granularity distributions of silver halide crystals. Broadly stated, multiple film layers are used with color film to separately capture and reproduce color information. The crystal granularity distributions further provide for capturing image details under different lighting and/or desired film exposure conditions or “speeds”. That is, larger or more “coarsely-grained” crystals require exposure to fewer photons of light during picture-taking to enable proper developing. Conversely, smaller or more “fine-grained” crystals that together encompass an area equal to that of a single larger grain will require more light during picture-taking to enable proper developing of each grain. Modem color film typically uses three layers or “emulsion coatings” for each color, each with a different granularity distribution, thereby enabling image detail to be recorded on the film at varying exposures for each color.
Unfortunately, while film can be viewed as recording all essential picture information, conventional chemical processing does not enable optimal reproduction of individual picture elements. More specifically, different picture elements such as highlights, midtones and shadows, while recorded on the same film layer, each require a different amount of development time to be optimally reproduced. Highlights, for example, might be optimally developed in one minute while midtones might require two minutes and shadows might require four minutes. However, chemical processing develops all picture elements at the same time. Thus, a compromise or approximation must be made as to an acceptable development time for accommodating all image elements.
Conventional techniques have been developed to improve the overall quality of a picture by taking into account varying developing time requirements. However, the resulting picture element handling capabilities are nevertheless subject to chemical processing limitations. For example, using monochrome film having multiple granularity distribution, and thus varying exposure-sensitivity layers, multiple exposures of a single still image can be recorded and later merged in a darkroom. However, such techniques still provide only gross adjustments to multiple picture element combinations, and require extensive effort and often specialized film and/or processing to do so.
Digital film processing (“DFP”) takes an entirely different approach. DFP provides for capturing raw image data directly from the film itself while it is being developed. Each channel (such as red, green, blue) of each element or pixel will typically be captured separately at multiple developing time increments or timed-scans. The pixel data of each timed-scan is then analyzed and/or manipulated to provide optimally an image that uses the appropriate exposure for each pixel.
FIG. 1
illustrates certain aspects of a DFP system taught in U.S. Pat. No. 5,519,510, and is taken from
FIG. 9
of that patent. As shown, the DFP system provides scanning a system for scanning an image recorded on film
101
directly from film
101
while film
101
is being developed. At multiple times during development of film
101
, infrared light sources
104
or light-source-array projects infrared light
11
and
12
at color film
101
and a portion of the image is captured (i.e. “timed-scanned”). Scanning a three-layer film
101
, for example, includes capturing reflected light
11
from portions (e.g. portion
11
a
) contained in the first or “front” film layer
111
and capturing reflected light
12
from portions (e.g. portion
12
a
) contained in a third or “back” film layer
113
. Portions (e.g. portion
11
a
) contained in the second (i.e. “middle” or “through”) layer are also captured by scanning transmitted light
11
passing through film
101
, from which scanned front layer and back layer values for corresponding front layer and back layer portion scans are then subtracted. This process is repeated for each portion, thereby producing front-layer, middle-layer and back-layer portion information for each portion position at each timed-scan during development of film
101
. Scanned portion information (or “pixel information”) is then processed as described hereinafter with reference to
FIGS. 2
a
-
3
.
As shown in
FIG. 2
a
, during picture-taking with camera
210
, a single picture
201
recorded onto film
101
or “exposed” film (
FIG. 1
) will typically include discernable picture elements such as highlights
211
a
, midtones
211
b
and shadows
211
c.
Turning to
FIG. 2
b
, exposed film
101
is then subjected to DFP processing. First, the above-mentioned timed-scans are taken using scanning system. During processing of film
101
, an early scan
202
a
(e.g. one minute) will best reveal pixels corresponding to highlights
211
a
, while midtones
211
b
and shadows
211
c
will be underdeveloped. A later scan (e.g. two minutes) will better reveal midtones
211
b
, while highlights
211
a
will become overdeveloped. Still later scans will better reveal shadows
211
c
at the expense of highlights
211
a
and midtones
211
b.
FIG. 3
illustrates that while several timed-scans are typically taken, not all of the scans obtained throughout film development are typically required. Rather, a sufficient number of scans are desirably taken such that the optimal exposure of selected picture elements in each film layer can be deduced by extrapolating from the limited number of scans actually taken. Individual scans can further be combined to reduce memory requirements. For example, scans
302
and
304
can be combined to produces scan
308
.
FIG. 3
also illustrates how the DFP system further provides an image-based processing/stitching system for processing and merging of groups of picture elements in order to form a completed processed image. Creating a merged image will be generally referred to herein as “stitching” regardless of the specific implementation utilized. As taught in the 510 patent, an approximation is preferably made as to the best exposure for different groups of picture elements utilized (i.e. in this case, highlight, midtone and shadow portions of the image stored on the film). Next, the different groups of picture elements can be combined by aligning, cutting and pasting them together
320
to yield the finished image
322
.
Unfortunately, if many timed-scans are needed, memory requirements become exorbitant. Accordingly, when fewer timed-scans are used, or when the actual time-scan time varies from the desired time-scan time, estimates of timed-scans which estimates were not reliable, needed to be made in order for the groups of picture elements to appear as a single continuous picture. Thus, adjustment approximations according to available scanned data were required, often yielding less-than-ideal results and often requiring significant computation.
U.S. Patent Application Ser. No. 60,075,562, filed Feb. 23, 1998, entitled “Parametric Image Stitching” teaches an alternative method of stitching, and is hereby expressly incorporated by reference. In that application, the tedium of pasting groups of picture elements is replaced by forming a mathematical model representing the recorded image and then resolving the model as a completed image. Modeling, in this instance, generally includes the ability to process stitch picture elements (e.g. pixels) without reference to the overall picture formed on the film. As described, received timed-scan pixel data is tagged with the time of capture (i.e. a relative and/or absolute indicator of the time of each scan during the course of film processing). Then, for each pixel at each time, regression parameters are c

Affiliated with

Also associated with

Rating

Log time processing and stitching 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 Log time processing and stitching system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Log time processing and stitching system will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3022747