 |
Description  |
|
|
BACKGROUND OF THE INVENTION
The invention generally relates to systems and methods for color correcting
video picture signals and for detecting scene changes during color
correction operations. More particularly, the invention pertains to
improved systems and methods for increasing the quality and speed of color
correction operations by enhancing the ability of color correction
equipment to determine when a new scene begins. This patent application
describes improvements upon the color correction systems and methods
disclosed in U.S. Pat. No. 4,096,523 (the "Rainbow" patent); No. 4,223,343
(the "Anamorphic" patent); No. 4,410,908 (the "Luminance" patent); No.
4,679,067; and No. 4,694,329; as well as those disclosed in copending,
commonly owned U.S. Pat. Appl. Ser. No. 807,815, entitled "Editing System
and Method"; Ser. No. 851,164, entitled "Color Correction System and
Method"; Ser. No. 942,901, entitled "Color Correction System and Method";
Ser. No. 943,218, entitled "Color Correction System and Method"; and Ser.
No. 943,298, entitled "Color Correction System and Method." The
disclosures of these patents and patent applications are hereby
incorporated herein by reference.
There is a continuing need to improve the efficiency, speed, and quality of
the color correction of video picture signals, especially in film-to-tape
and tape-to-tape transfers, and particularly in scene-by-scene color
correction. For instance, there is a need for equipment that more
accurately senses new scenes in a motion picture film or a videotape that
is being color corrected. Furthermore, there is a need to prevent the
physical degradation of motion picture film and videotape caused by
scratching due to the back-and-forth movement necessary to find the
beginning of a scene. Moreover, there is a need to reduce the time an
operator spends hunting for the start of a scene.
An accurate scene-change detector is especially important when a videotape
is being color corrected, since the image may change at the video field
rate of 60 hertz. By contrast, when a motion picture film is being color
corrected, the image may change at the frame rate of 24 hertz. Hence,
finding the start of a new scene on a videotape may be very difficult and
time-consuming to accomplish manually inasmuch as more images appear
during a given period than with a film.
A scene-change detector or analyzer is advantageously used with a color
corrector, as indicated in an article entitled "The Pre-Programming of
Film-Scanner Controls," by D. J. M. Kitson, A. B. Palmer, R. H. Spencer,
J. R. Sanders, and M. Weston, which was published in E.B.U. Review, No.
134, August 1972, on pages 156-162, and an article entitled "Preprogrammed
and Automatic Color Correction for Telecine," by D. J. M. Kitson, J. R.
Sanders, R. H. Spencer, and D. T. Wright, which was published in the
Journal of the SMPTE, Volume 83, August 1974, on pages 633-639. There is a
need for improvement of scene-change detectors or analyzers.
OBJECTS OF THE INVENTION
Accordingly, an object of the invention is to satisfy the above needs and
provide a system and method for color correcting video picture signals
with increased efficiency, speed, and quality.
Another object of the invention is to provide a system and a method for
improving the accuracy with which the start of a new scene may be
ascertained.
An additional object of the invention is to provide a signal processing
device and a method for reducing the number of new scenes missed by a
scene-change detector.
A further object of the invention is to provide a system and a method for
preventing the physical degradation, e.g., scratching, of motion picture
film and videotape caused by jogging the recording medium back and forth
when hunting for the start of a new scene.
Yet another object of the invention is to provide an apparatus with
improved signal processing circuits and a method with improved signal
processing techniques.
Still another object of the invention is to provide improved devices and
techniques for analyzing various video signal parameters in order to
ascertain when a new scene begins.
Another object of the invention is to provide a scene-change detector and
corresponding signal processing method that accurately analyze even
low-level video signals to sense the start of a new scene.
SUMMARY OF THE INVENTION
The invention satisfies the needs identified above and meets the foregoing
objects by providing a system which is better able to sense scene changes
in a succession of video picture signals. In accordance with one aspect of
the invention, a color corrector includes a scene change detection module
which processes video picture signals to detect when the corresponding
images start a new scene. Specifically, the scene change detection module
includes circuits for analyzing a change in each of at least two different
parameters of the video picture signals. Each change is independently
compared with a predetermined standard, and the scene change detection
module generates a change detect signal when the change in at least one of
the two parameters satisfies the predetermined standard. Preferably, an
area discrimination circuit is provided for the scene change detection
module. Such an area discrimination circuit may permit the operator to
selectively control the portion of the video picture in which the video
picture signals are analyzed.
The scene change detection module advantageously analyzes a video signal
parameter that is indicative of the color content of the picture as well
as a video signal parameter that corresponds to signals in a preselected
frequency range. The preselected frequency range is preferably below the
range of frequencies for chrominance signals, and may be between about 1.5
megahertz and about 2.5 megahertz.
In accordance with another aspect of the invention, a scene change analyzer
includes a bandpass filter which transmits video signals within a
preselected pass band. The scene change analyzer also includes scene
sensing circuits which are responsive to the video signals transmitted by
the bandpass filter. Such scene sensing circuits may be provided with
individual circuits or a programmable device for determining a first
average level of the transmitted video signals in at least a portion of a
first video field and for determining a second average level of the
transmitted video signals in at least a portion of a second video field.
The difference between the average levels is then compared with a
predetermined standard in order to test whether a new scene has started.
If the predetermined standard is satisfied, the scene change analyzer
produces an output signal indicative of a scene change. Preferably, the
pass band of the bandpass filter is centered at approximately 2.0
megahertz and has a width of about 1.0 megahertz around the center
frequency.
The features of the invention each improve the ability of the equipment to
detect the start of a new scene and increase the efficiency of the color
correction process. Such features enable an operator to color correct a
motion picture film or a videotape more efficiently, thereby reducing the
cost of the color correction procedure.
BRIEF DESCRIPTION OF THE DRAWINGS
The above and other objects, features, and advantages of the present
invention will become apparent upon consideration of the following
detailed description of illustrative embodiments thereof, especially when
taken in conjunction with the accompanying drawings, wherein:
FIG. 1 is a diagrammatic illustration of a color correction system with a
scene acquisition and sensing module according to the invention;
FIG. 2 is a block diagram of a scene acquisition and sensing module
according to the invention;
FIGS. 3A and 3B are enlarged views of the monitor shown in FIG. 2 and
illustrate one type of display that may be employed with a scene
acquisition and sensing module according to the invention; and
FIG. 4 is a flowchart of a routine that may be utilized to analyze video
signals and sense the start of a new scene.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
General Description
FIG. 1 shows a color correction system 10 which includes a color corrector
11 having a front panel 12. Portions of the front panel 12 are illustrated
in greater detail in FIGS. 2-4 of application Ser. No. 851,164;
application Ser. No. 942,901; application Ser. No. 943,218; and
application Ser. No. 943,298. The front panel 12 has a set of variable
vector controls 14 and a set of six vector controls 16. The six vector
controls 16 function as outlined in the Rainbow and Luminance patents,
which are mentioned above.
Referring now to the lower left-hand portion of FIG. 1, the front panel 12
includes a set of color balance controls 18 and "window" controls 20. The
"window" controls 20 are described and depicted in greater detail in U.S.
Pat. No. 4,679,067 as well as in U.S. Pat. No. 4,694,329. The front panel
12 additionally includes video signal source controls 22. A telecine or a
videotape recorder/reproducer may be employed as the video signal source.
The video signal source controls 22 may adjust parameters such as the PEC
gain and negative gain for each of the red, green, and blue channels.
Moreover, the video signal source controls may adjust other parameters,
for instance, the horizontal pan, the vertical pan, the zoom, and the
contours. Each of the controls in the sets of controls 14, 16, 18, and 22
includes a control knob which is coupled to a shaft-position encoder, as
discussed in U.S. Pat. No. 4,679,067 and U.S. Pat. No. 4,694,329.
The right side of the front panel 12 includes pushbuttons and displays.
Specifically, this portion of the front panel includes two rows of
pushbuttons 24, which are shown in greater detail in FIG. 4 of the
above-identified patent applications, and three rows of pushbuttons 26,
which are shown in greater detail in FIG. 3 of the above-identified patent
applications. The functions of many of these pushbuttons are explained in
the Rainbow and Luminance patents. A display 28 shows the scene number for
the color corrections stored in the A buffer and the B buffer. Moreover,
the display 28 shows the scene number for the current scene.
Still referring to FIG. 1, a keypad 30 and a display 32 are used to recall
the color corrections for a particular scene and apply them to the present
scene. For example, if the operator wanted to use the color corrections
for previous scene number 1,234 and apply them to the current scene, the
operator would press the "call" pushbutton in the upper one of the rows 24
and then the buttons 1, 2, 3, and 4 of the keypad 30 in this sequence in
order to recall the desired color corrections.
Also shown in FIG. 1 is an array 34 of pushbuttons and a row of pushbuttons
36 for use in the "Call-A-Picture" feature of the color correction system.
The operation of the "Call-A-Picture" feature is described in application
Ser. No. 943,298. The upper right portion of the front panel 12 depicted
in FIG. 1 has waveform pushbuttons and indicators 38 for selecting various
waveforms for viewing on an oscilloscope (not shown) as well as monitor
selector pushbuttons and indicators 40 for selecting various signals for
monitoring.
As illustrated in FIG. 1, the system 10 has a computer 42, which is
connected to each of the color corrector 11, a video signal source 44, a
videotape recorder 46, and a video memory 48. The video signal source 44
may be a film chain or telecine, a videotape player, or the like. The
video signal source 44 produces video signals from the associated image
recording medium. These video signals are delivered to the color corrector
11 so that they can be corrected. The color corrector 11 provides color
corrections for the video signals from the video signal source 44 under
the direction of the operator and the computer 42, and it produces color
corrected video signals. The color corrected video signals are sent to a
main monitor 50, and, at the appropriate time, to the videotape recorder
46. The operator may observe the effect of the color corrections on the
video signals by looking at the video picture on the main monitor 50. The
videotape recorder 46 records the color corrected video signals on a
videotape 54, usually during a second run after color corrections have
been made during a first run, thereby producing a color corrected
videotape.
The main monitor 50 is shown with windows W1 and W2. One use of the windows
W1 and W2, which are movable in size and/or position, is described in the
above-identified patent applications. Other uses of the windows are
discussed in U.S. Pat. No. 4,679,067 and U.S. Pat. No. 4,694,329.
An auxiliary monitor 52 is connected to the computer 42. The auxiliary
monitor 52 displays a plurality of video pictures, such as the video
pictures 56a-56d. The auxiliary monitor 52 and the video memory 48 are
employed to implement the "Call-A-Picture" feature of the color correction
system.
FIG. 1 shows a scene acquisition and sensing module 60 according to the
invention, which module is illustrated in greater detail in FIG. 2. The
module 60 is connected to receive output signals from the color corrector
11. Namely, the color corrector 11 supplies color corrected video signals
to the module 60. The module 60 processes these video signals to detect
the start of a new scene, as explained below. The module 60 delivers
output signals to a supplemental monitor 62.
The supplemental monitor 62 displays video pictures formed by the output
signals from the color corrector 11. The supplemental monitor 62 also
provides the operator with a bar graph display 70, which is discussed
below in connection with FIGS. 3A and 3B, and it shows a window W3. The
window W3 is generated by the module 60 under the control of the operator,
and the window W3 is independent of the windows W1 and W2, which are
displayed on the main monitor 50. The window W3 denotes the portion of the
video picture in which the module 60 analyzes video signals to detect the
beginning of a new scene. The operator may adjust the size and/or position
of the window W3 as desired.
The module 60 is connected to send signals to and receive signals from the
computer 42. The module 60 receives frame pulse and direction signals from
the computer 42. The reason that these signals are delivered to the module
60 is explained below as part of the discussion of the flowchart of FIG.
4. The module 60 transmits a change detect signal to the computer 42 when
it locates the start of a new scene, and the computer 42 then sends an
appropriate signal to the color corrector 11.
FIG. 1 depicts the module 60 connected to the supplemental monitor 62,
which may be a black-and-white monitor in order to reduce the cost of the
system. However, the module 60 may be connected to the main monitor 50. If
so, the module 60 is preferably connected through a switching circuit that
will enable the operator to selectively control the presence of the bar
graph display 70 and the window W3 on the main monitor 50. At times, the
bar graph display 70 and the window W3 may be distracting to the operator
when the operator is attempting to color correct the video pictures
appearing on the main monitor 50. Accordingly, the presence of the bar
graph display 70 and the window W3 on the main monitor 50 is
advantageously controllable by the operator with a suitable switching
circuit.
FIG. 1 shows the module 60 connected at the output of the color corrector
11. The module 60 may alternatively be connected at another point in the
system. For instance, the module 60 may be located to receive uncorrected
video signals from the video signal source 44.
Either composite video signals or component video signals may be delivered
to the module 60. For ease of explanation, a module that processes
composite video signals will be described below. However, a person having
ordinary skill in the art will readily recognize how such a module may be
modified to process component video signals.
Scene Acquisition and Sensing Module
FIG. 2 better illustrates a scene acquisition and sensing module 60
according to the invention. As shown in the upper left-hand portion of
FIG. 2, composite video signals are delivered to a conventional color
decoder 80. The color decoder 80 forms a luminance signal ("Y"), a signal
representing the difference between the red and luminance signals ("R-Y"),
and a signal representing the difference between the blue and luminance
signals ("B-Y") from the input composite video signals. The color decoder
80 also derives horizontal drive signals and vertical drive signals from
the composite video signals.
The color decoder 80 sends the Y, R-Y, and B-Y signals to a quad integrator
unit 82. The color decoder 80 also sends the Y signal to a bandpass filter
84, which transmits its output signal to the quad integrator unit 82. A
main processing unit 86 receives the horizontal drive signals and the
vertical drive signals from the color decoder 80.
Preferably, the bandpass filter 84 has a center frequency of approximately
2.0 megahertz and a pass band of about .+-.0.5 megahertz around the center
frequency. This range of frequencies corresponds to signals indicative of
the sharpness or detail of images in a video picture. Accordingly, the
output signals from the bandpass filter 84 are referred to as the detail
signals in the following description. These detail signals may be
advantageously employed to sense the start of a new scene, either alone or
in combination with color-indicative signals, e.g., R-Y and B-Y signals,
and/or the luminance signal, as explained further below.
The quad integrator unit 82 operates to sample the detail, Y, R-Y, and B-Y
video signals in each field. The quad integrator unit 82 then supplies the
sampled signals to the main processing unit 86, which analyzes the sampled
signals from successive fields to determine whether a new scene has
started. In order to accomplish its signal sampling function, the quad
integrator unit 82 includes four integrator circuits 88, 90, 92, and 94.
Each of these integrator circuits receives a different video signal at its
input. Specifically, the output signals from the bandpass filter 84, i.e.,
the detail signals, are supplied to the integrator circuit 88, while the
Y, R-Y, and B-Y signals are sent to the integrator circuits 90, 92, and
94, respectively. Thus, each integrator circuit independently samples the
associated video signals.
Each of the integrator circuits 88, 90, 92, and 94 may comprise the
integrator 304, the switching circuits 312 and 314, the integrator 316,
and the buffer amplifier 318 which are shown in FIG. 9 of U.S. Pat. No.
4,694,329. Such integrator circuits determine the average level of the
associated video signal in the sampled picture area on a field-by-field
basis.
The main processing unit 86 delivers window signals to the quad integrator
unit 82. The window signals are used to select the portion of the video
picture in which the integrator circuits 88, 90, 92, and 94 are operative.
In other words, the window signals from the main processing unit 86
control the integrator circuits so that they only sample video signals in
a limited area of the video picture. This type of signal sampling for
purposes of scene change detection is described in U.S. Pat. No.
4,694,329. As discussed in that patent, the size and the location of the
area may be selectively controlled by the operator, who adjusts the area
to obtain optimal performance.
For each field, the quad integrator unit 82 supplies four samples to the
main processing unit 86. In particular, these are samples of the detail,
Y, R-Y, and B-Y signals, and they are sent to a bank of analog-to-digital
converters 96 in the main processing unit 86. The bank 96 includes an
analog-to-digital converter for each of the four field samples. The bank
96 also includes an analog-to-digital converter which receives a signal
from a sensitivity potentiometer 98.
The sensitivity potentiometer 98 is used to adjust the sensitivity, or
threshold level, of the scene acquisition and sensing module 60. The
output signal of the sensitivity potentiometer 98 corresponds to the
threshold level signal shown in FIG. 9 of U.S. Pat. No. 4,694,329. A
relatively high output signal from the potentiometer 98 results in a
relatively high threshold level, and a relatively large change in the
sampled signals is needed before a change detect signal is generated.
Conversely, a comparatively low output signal from the potentiometer 98
results in a comparatively low threshold level, and a comparatively small
change in the sampled signals will produce a change detect signal.
The module 60 independently analyzes each of the detail, Y, R-Y, and B-Y
signals to detect the start of a new scene. That is, a change detect
signal is generated when any one of these video parameters changes
sufficiently so that the change exceeds the threshold level. Since four
different parameters are being processed simultaneously and since a
sufficient change in any one of these parameters may produce a change
detect signal, the module 60 detects new scenes with greater accuracy than
conventional devices. Fewer scene changes are missed.
Although a single sensitivity potentiometer is shown, a sensitivity
potentiometer for each of the sampled video parameters may be provided.
With this alternative arrangement, the threshold level for each of the
sampled video parameters may be adjusted independently of the others.
The threshold level may be adjusted by a knob (not shown) on the front of
the module 60, which knob is connected to the sensitivity potentiometer
98. In addition, the threshold level may be set and reset with the
controls of the color corrector 11. For example, the "window" controls 20
may be operated to set or reset the threshold level by pressing the "size"
pushbutton and holding it down and then by pressing the "arrow up"
pushbutton or the "arrow down" pushbutton. The "arrow up" pushbutton is
actuated to increase the threshold level, while the "arrow down"
pushbutton is actuated to decrease the threshold level. These pushbuttons
are shown in FIG. 1 of U.S. Pat. No. 4,679,067 and U.S. Pat. No.
4,694,329.
The analog-to-digital converters in the bank 96 supply digital
representations of the field samples from the quad integrator unit 82 to
the microprocessor 100. Furthermore, one of the analog-to-digital
converters in the bank 96 delivers a digital representation of the
threshold level from the sensitivity potentiometer 98 to the
microprocessor 100. As explained in connection with the flowchart shown in
FIG. 4, the microprocessor 100 analyzes the samples from various fields
for each of the detail, Y, R-Y, and B-Y channels in order to detect the
start of scene. When a scene change is sensed in any of the channels, the
microprocessor 100 supplies a change detect signal to the computer 42
(FIG. 1) through an output port 102.
To accomplish its scene sensing function, the microprocessor 100 receives
vertical drive signals from the color decoder 80 and frame pulse and
direction signals from the computer 42 (FIG. 1). The signals from the
computer 42 are delivered to the microprocessor 100 through an input port
104.
In addition, the microprocessor 100 receives input signals from the toggle
switches 106, 108, and 110 through an input port 112 along with signals
from the DIP (dual in-line pin) switches 114 and 116. The toggle switches
106, 108, and 110 together with the DIP switches 114 and 116 are used to
configure the equipment to the particular needs or desires of the user.
For instance, the individual switches constituting each of the DIP
switches 114 and 116 may be set to establish whether the equipment
operates with or without a certain feature. The individual switches
forming a DIP switch may control such options as whether the equipment is
used to detect scene changes for motion picture film or videotape, whether
a linear lookup table or a logarithmic lookup table is employed for the
field samples (see the discussion of the flowchart of FIG. 4), and whether
all or only some of the four channels are analyzed by the microprocessor
100.
The "A/B mode" toggle switch 106 determines whether the microprocessor
reads the individual switches in DIP switch 114 or the individual switches
in DIP switch 116. In other words, one of the DIP switches 114 and 116
contains the presets for the A mode of the equipment, while the other of
the DIP switches 114 and 116 contains the presets for the B mode of the
equipment. The function of the "reject on/off" toggle switch 108 will be
discussed below in connection with the description of the flowchart of
FIG. 4. Briefly, however, this switch determines whether a single
greater-than-threshold-level difference or whether consecutive
greater-than-threshold-level differences are necessary to produce a change
detect signal. The "PAL/NTSC" toggle switch 110 is operated to inform the
microprocessor 100 of the format of the video signals being analyzed.
As shown in FIG. 2, the microprocessor 100 supplies control signals to a
window generator 118 and a bar display generator 120. The window generator
118 also receives horizontal drive signals and vertical drive signals from
the color decoder 80. The window generator 118 is used to produce the
window signals that control the size and position of the area of the video
picture in which the quad integrator unit 82 samples the detail, Y, R-Y,
and B-Y signals. The window generator 118 also produces window outline
signals for display on a monitor. The bar display generator 120 produces
bar formation signals for display on a monitor. As noted previously during
the description of FIG. 1, this monitor may be the main monitor 50 or the
supplemental monitor 62.
The window generator 118 may be identical to the window generator 310
illustrated in FIG. 9 of U.S. Pat. No. 4,694,329. The window generator 118
may comprise four programmable counters, each of which receives its count
signal from the microprocessor 100. Such an arrangement is illustrated in
FIG. 5 of U.S. Pat. No. 4,679,067 and U.S. Pat. No. 4,694,329. The four
programmable counters determine the horizontal width and the vertical
height of the window. The window outline signals may be formed by one-shot
circuits which produce pulse signals when the programmable counters change
state.
The bar display generator 120 may include programmable counters, too. The
microprocessor 100 delivers signals representing the magnitude of the
threshold level, or sensitivity setting, and the difference between
successive field samples in various channels to the programmable counters.
The programmable counters then operate to generate output pulse signals,
where the width of each output pulse corresponds to the magnitude of the
associated parameter.
The bar formation signals and the window outline signals are supplied to an
amplifier 122 together with composite video signals from the input of the
module 60. The amplifier 122 combines the composite video signals and the
display signals and delivers its output signal to a monitor 124, e.g., the
main monitor 50 or the supplemental monitor 62 of FIG. 1. The operator,
therefore, may observe the video picture along with the bar graph display
70 and the window W3 on the monitor 124. The window W3 corresponds to the
area of the video picture in which the quad integrator unit 82 samples the
detail, Y, R-Y, and B-Y signals. As explained above, the size and position
of the window W3 may be selectively changed by the operator. The bar graph
display 70 is illustrated in greater detail in FIGS. 3A and 3B.
Referring now to FIG. 3A, the bar graph display 70 on the monitor 124 is
formed from four bars 126, 128, 130, and 132. The bar 126 denotes the
current threshold level for the scene acquisition and sensing module 60. A
shorter sensitivity bar 126 denotes a lower threshold level, while a
longer sensitivity bar 126 denotes a higher threshold level.
The bars 128, 130, and 132 indicate the frame-by-frame or field-by-field
difference in the Y, R-Y, and B-Y signals, respectively. While display
bars for the Y, R-Y, and B-Y signals are illustrated, the module 60 may
generate display bars for additional or alternative signals. For instance,
a bar designating the frame-by-frame or field-by-field difference in the
detail signal may be displayed, as may a bar indicative of the
frame-by-frame or field-by-field difference in the absolute value of the
[(R-Y)-(B-Y)] signal.
FIG. 3A depicts a typical display when no new scene has been detected. Each
of the bars 128, 130, and 132 is shorter than the sensitivity bar 126,
which means that none of the Y, R-Y, and B-Y signals has changed
sufficiently to exceed the threshold level. A dashed line 134 is drawn in
FIG. 3A so that the length of the sensitivity bar 126 may be easily
compared to the lengths of the bars 128, 130, and 132. FIG. 3B, on the
other hand, shows a typical display when a new scene has been detected.
Specifically, the Y difference bar 128 extends beyond the sensitivit | | |
