JPH1169350A - Digital moving image coding system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a digital moving image coding system which can effectively code a moving image and also can improve the picture quality of the image by performing the conversion of the coding time resolution so as to evade occurrence of an image where the cinema and NTSC images are mixed together. SOLUTION: When the NTSC signals of 30 frames/s, for example, are inputted as input data 151, a cinema detection part 101 decides whether a cinema sequence that is produced at a cinema frame rate is available. The detected sequence is notified to a time resolution conversion part 102 by a control signal 152. Then the part 102 converts 30 frames/s into 24 frames/s, i.e., converts 5 frames into 4 frames. In other words, it's possible to effectively code the cinema sequence with 24 frames/s of the original picture and to obtain an image of high picture quality by performing the time resolution conversion from 30 frames/s into 24 frames/s via both parts 101 and 102.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a digital moving picture coding system such as a digital video transmitting apparatus, a digital CATV, and a digital broadcasting system.

[0002]

2. Description of the Related Art FIG. 17 shows "Grand Allianc".
e HDTV System Specificati
on April 14, 1994 ", which is a time resolution conversion method of a conventional digital image coding method.

A conventional digital image coding system will be described with reference to FIG. On the encoding side, NTSC (Nati
onal Television System Co
When converting a cinema image input at a frame rate into a cinema frame rate, for example, when the NTSC frame rate is 30 frames / s and the cinema frame rate is 24 frames / s, the cinema image is converted from 5 frames to 4 frames. I do. Here, the frame is 1
One image and two images are shown, and are composed of an odd field and an even field for display on a television monitor. In that case, the first frame of pack 4 comprises the first frame of pack 5 (or the odd field of the first frame of pack 4 comprises the odd field of the first frame of pack 5 and the odd field of the second frame. ,
The even field of the first frame of pack 4 is composed of the even field of the first frame of pack 5, and the second frame of pack 4 is composed of the odd field of the third frame of pack 5 and the even frame of the second frame. The third frame of pack 4 is composed of the odd field of the fourth frame of pack 5 and the even frame of the third frame of pack 5 (or the odd field of the third frame of pack 4 is the odd field of the fourth frame of pack 5). And the even field of the third frame of the pack 4 is composed of the even field of the third frame and the even field of the fourth frame of the pack 4, and the fourth frame of the pack 4 is formed from the fifth field of the pack 5. Configure and perform time resolution conversion.

On the decoding side, when converting from a cinema frame rate to an NTSC frame rate, the first frame of pack 5 is the first frame of pack 4 and the second frame of pack 5 is the first frame of pack 4. From the odd field of one frame and the even field of the second frame, the third frame of pack 5 is obtained from the odd field of the second frame of pack 4 and the even field of the third frame, and the fourth and fifth frames of pack 5 are obtained. Each frame is composed of the third frame and the fourth frame of the pack 4, and the encoding side adds to the transmission data so that the time resolution conversion is performed on the decoding side. In this case, the conversion is performed irrespective of the structure of the pack 5 sequence.
May be a mixed sequence of a cinema image and an NTSC image, or may be a scene change in a cinema sequence. In this case, the coding efficiency will be degraded.

[0005]

As described above, in the conventional digital image coding method, a cinema sequence and an NTSC sequence are mixed in a unit for performing conversion,
Even when there is a scene change in the cinema sequence, constant conversion is performed, so that the encoding must be performed due to the discontinuity of the image, resulting in a problem that the image quality deteriorates.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and is intended for a case where a cinema image having a cinema frame rate and an NTSC image such as a commercial having an NTSC frame rate exist in a continuous sequence. At the connection point between the cinema image and the NTSC image, the encoding side time resolution conversion is performed so as to avoid an image in which the cinema image and the NTSC image are mixed, thereby efficiently encoding the image and improving the image quality. Even if there is a scene change in the rate cinema image, encoding is performed efficiently by performing encoding-side time resolution conversion to avoid images in which images before and after the scene change are mixed, thereby improving image quality. It is intended to obtain a digital video coding system which can be performed.

[0007]

In the digital moving picture coding system according to the present invention, the input data is a normal N
A cinema detecting means for determining whether the sequence is a TSC sequence or a cinema sequence created at a cinema frame rate; and removing a redundant field from the sequence determined to be a cinema sequence by the cinema detecting means, and converting the cinema sequence from the NTSC frame rate of input data. A time resolution conversion means for performing a time resolution conversion to a frame rate, and a time resolution conversion for optimizing the coding efficiency in a boundary portion between a normal NTSC sequence and a cinema sequence or a scene change portion in a cinema sequence; When decoding a cinema sequence that has been converted to a cinema frame rate and coded, the decoding information is displayed at the NTSC frame rate. It is obtained by a coding means for adding.

In the resolution conversion on the encoding side, the time resolution conversion means performs time conversion on the input data of the NTSC frame rate for each specific unit (hereinafter referred to as pack 5) (the unit after time conversion of pack 5). In the case where a cinema sequence and a normal NTSC sequence are included in the pack 5, the pack 5 is converted from the pack 5 to the pack 4 by performing specific processing so that the coding efficiency is not deteriorated. It is.

In the case where the pack 5 is composed of 5 frames and the pack 4 is composed of 4 frames, the first frame in the pack 5 is a cinema sequence and the second and third , When the fourth and fifth frames are the normal NTSC sequence, the first frame of pack 5 to the first frame of pack 4, the second frame of pack 5 to the second frame of pack 4, and the fourth frame of pack 5 The third frame of pack 4 is composed of the even field of three frames and the odd field of the fourth frame,
The fifth frame of the pack 5 constitutes the fourth frame of the pack 4.

When the pack 5 is composed of 5 frames and the pack 4 is composed of 4 frames, the first and second frames in the pack 5 are cinema sequences, , When the fourth and fifth frames are normal NTSC sequences,
The pack 4 is constituted by deleting the frame.

Further, the time resolution converting means, when the pack 5 is composed of 5 frames and the pack 4 is composed of 4 frames, the first, second and third frames in the pack 5 are cinema sequences, When the fourth and fifth frames are normal NTSC sequences, the first, second, and third frames of pack 5 are converted to the first and second frames of pack 4, and the fourth and fifth frames of pack 5 are converted. The conversion into the third and fourth frames of the pack 4 is performed.

[0012] When decoding the encoded cinema frame rate data, the encoding means converts the cinema frame rate to the NTSC frame rate on the decoding side. 5 are converted from the first frame of pack 4 and the second frame of pack 4, respectively, and the third frame of pack 5 is the odd field of the second frame of pack 4 and the third frame of pack 4. The fourth and fifth frames of pack 5 are created from the even field of the frame, and the information created from the third and fourth frames of pack 4 is added to the encoded data.

When decoding the encoded cinema frame rate data, the encoding means converts the cinema frame rate to the NTSC frame rate on the decoding side. The first frame of pack 5 is converted to the first frame of pack 4, the second frame of pack 5 is created from the odd field of the first frame of pack 4 and the even field of the second frame of pack 4, The third, fourth, and fifth frames add information created from the second, third, and fourth frames of pack 4 to the encoded data.

In the digital video coding system according to the present invention, a cinema detecting means for determining whether input data is a normal NTSC sequence or a cinema sequence created at a cinema frame rate, and the cinema detecting means comprises a cinema detecting means. Redundant fields are removed from the sequence determined to be a sequence, and NT
Encoding-side time resolution conversion means for performing time resolution conversion from the SC frame rate to the cinema frame rate; converted scene change detection means for performing scene change detection on the sequence after the time resolution conversion; Encoding means for adding instruction information of time resolution conversion for displaying an image decoded at the time resolution of the cinema sequence in a normal NTSC sequence to the encoded data.

[0015] When the scene change detecting means detects that a scene change has occurred between the first frame and the second frame in the pack 4, the encoding means performs decoding. When performing time resolution conversion from the pack 4 of 4 frames to the pack 5 of 5 frames of NTSC, the first frame of pack 5 is from the first frame of pack 4, and the second frame of pack 5 is the second frame of pack 4. , The third frame of pack 5 is from the odd field of the second frame of pack 4 and the even field of the third frame, and the fourth frame of pack 5 is
From the third frame, the fifth frame of the pack 5 adds information of an instruction to the encoded data so as to be created from the fourth frame of the pack 4.

When the scene change detecting means detects that a scene change has occurred between the second frame and the third frame in the pack 4, the encoding means performs decoding. When performing time resolution conversion from the pack 4 of 4 frames to the pack 5 of 5 frames of NTSC, the first frame of the pack 5 is from the first frame of the pack 4, and the second frame of the pack 5 is the first frame of the pack 4. From the odd field of the second frame and the even field of the second frame, the third frame of the pack 5
The instruction information is added to the encoded data so that the fourth frame of pack 5 is created from the third frame of pack 4, and the fifth frame of pack 5 is created from the fourth frame of pack 4.

Further, in the digital video coding system according to the present invention, a cinema detecting means for determining whether the input data is a normal NTSC sequence or a cinema sequence created at a cinema frame rate, and the cinema detecting means Redundant fields are removed from the sequence determined to be a sequence, and N
Encoding-side time resolution conversion means for performing time resolution conversion from a TSC frame rate to a cinema frame rate; pre-conversion scene change detection means for performing scene change detection on a sequence before time resolution conversion; Encoding means for adding an instruction of time resolution conversion for displaying an image decoded at a cinema frame rate at an NTSC frame rate to encoded data.

Further, the encoding-side time resolution converting means includes N
If the TSC frame rate is 30 frames / s and the cinema frame rate is 24 frames / s, pack 5
Is composed of 5 frames, pack 4 is composed of 4 frames, and if a scene change is detected between the first frame and the second frame in pack 5, the pack 5
, The first frame of the pack 4 is formed from the first frame of the pack 5, and the second
The third, fourth, and fifth frames constitute the second, third, and fourth frames of the pack 4.

The encoding-side time resolution converting means is arranged so that the pack 5 is composed of 5 frames, the pack 4 is composed of 4 frames, and the second frame and the third
When a scene change is detected between frames, when converting from pack 5 to pack 4, the second frame of pack 5 is deleted to form pack 4.

The encoding-side time resolution converting means is arranged so that the pack 5 is composed of 5 frames, the pack 4 is composed of 4 frames, and the third frame and the fourth
When a scene change is detected between frames, when converting from pack 5 to pack 4, the first and second pack 5
And the third frame are converted into the first and second frames of the pack 4, and the fourth and fifth frames of the pack 5 are converted into the third and fourth frames of the pack 4.

When decoding the encoded cinema frame rate data, the encoding means converts the pack 4 to pack 5 when converting the cinema frame rate to the NTSC frame rate on the decoding side. The first and second frames of pack 5 are converted from the first and second frames of pack 4, and the third frame of pack 5 is obtained from the odd field of the second frame of pack 4 and the even field of the third frame of pack 4. The fourth and fifth frames of pack 5 are for adding instruction information to the encoded data so as to be created from the third and fourth frames of pack 4.

Further, the encoding means, when decoding the encoded cinema frame rate data, when converting the cinema frame rate to the NTSC frame rate on the decoding side, the conversion from pack 4 to pack 5 The first frame of pack 5 is converted from the first frame of pack 4, and the second frame of pack 5 is created from the odd field of the first frame of pack 4 and the even field of the second frame of pack 4. The third, fourth, and fifth frames add instruction information to the encoded data to be created from the second, third, and fourth frames of pack 4.

[0023]

BEST MODE FOR CARRYING OUT THE INVENTION

Embodiment 1 FIG. Hereinafter, a digital moving picture coding method according to the present invention will be described.

FIG. 1 is a schematic block diagram showing a digital video coding system according to a first embodiment of the present invention.
FIG. 5 is a diagram showing the resolution conversion of the digital video coding system, and FIGS. 6 and 7 are diagrams showing the resolution conversion on the decoding side of the digital video coding system. In FIG. 1, 1
01 is a cinema detector, 102 is a time resolution converter, 10
3 is an encoding unit, 151 is input data, 152 is a control signal for instructing time resolution conversion, 153 is data after conversion,
Reference numeral 154 denotes encoded data.

Next, the operation will be described with reference to FIGS. When, for example, an NTSC signal of 30 frames / s is input as the input data 151, the cinema detection unit 101 determines whether or not the input data 151 is originally a cinema sequence created at a cinema frame rate. For example, a cinema sequence originally created at a cinema frame rate of 24 frames / s is converted from 24 frames / s to 30 frames / s. Is detected as a cinema sequence. The detected sequence is output to the time resolution conversion unit 1
02 is notified by the control signal 152. The time resolution conversion unit 102 converts the data into a pack 5 as shown in the time resolution conversion (cinema frame rate → NTSC frame rate) on the decoding side in FIG.
Re-conversion is performed as shown in the time resolution conversion (NTSC frame rate → cinema frame rate) on the encoding side in FIG. That is, 30 frames / s to 24 frames / s
Since it is converted to s, the conversion is from 5 frames to 4 frames. In a normal cinema sequence, the conversion shown in FIG. 17 may be performed. However, a cinema sequence and an NTSC sequence are mixed in five frames to be converted due to commercial insertion or the like, and a scene change is also included in the cinema sequence. In such a case, the coding efficiency is deteriorated by the reconversion in the normal method. Therefore, as shown in FIG. 2, when re-converting from 5 frames to 4 frames, re-conversion is performed in a manner that does not decrease the coding efficiency. FIGS. 3, 4 and 5 show specific examples of the resolution conversion shown in FIG.

In the case of FIG. 3, when performing conversion from five frames (pack 5) to four frames (pack 4), a cinema image and an NTSC image are mixed in the five frames, and the second frame and the third frame of the pack 5 are mixed. If the boundary is between frames, the second frame of pack 5 is dropped and conversion to pack 4 is performed. As a result, in the case of the conventional re-conversion method, the second frame of the pack 4 becomes a mixed frame of a cinema image and an NTSC image, and in the case of encoding, the efficiency is extremely deteriorated. The first frame of No. 4 is a cinema image, and the second, third and fourth frames are NTSC images, and the cinema image and the NTSC image can be distinguished without any problem.

In the case of FIG. 4, when performing conversion from five frames (pack 5) to four frames (pack 4), a cinema image and an NTSC image are mixed in the five frames, and the third frame and the fourth frame of the pack 5 are mixed. If the boundaries are between frames, the first frame of pack 4 is from the first frame of pack 5, the second frame of pack 4 is from the odd field of the third frame of pack 5, and the even field of the second frame of pack 5. , And the third and fourth frames of pack 4 perform the conversion from the fourth and fifth frames of pack 5, respectively. As a result, in the case of the conventional re-conversion method, the third frame of the pack 4 becomes a mixed frame of a cinema image and an NTSC image, and in the case of encoding, the efficiency is extremely deteriorated. 4 is a cinema image, and the third and fourth frames are NT.
The image becomes an SC image, and a cinema image and an NTSC image can be distinguished without any problem, and encoding can be improved.

In the case of FIG. 5, when converting from 5 frames (pack 5) to 4 frames (pack 4), a cinema image and an NTSC image are mixed in the 5 frames, and the first frame and the second frame of the pack 5 are mixed. In the case of a boundary between the frames, the first and second frames of pack 4 are from the first and second frames of pack 5, and the third frame of pack 4 is the odd field of the fourth frame of pack 5 and the first and second frames of pack 5. 3
From the even field of the frame, the fourth frame of pack 4 performs the conversion from the fifth frame of pack 5, respectively. As a result, in the case of the conventional re-conversion method, the third frame of the pack 4 becomes a mixed frame of a cinema image and an NTSC image, and in the case of encoding, the efficiency is extremely deteriorated. The first frame No. 4 is a cinema image, and the second, third and fourth frames are NTSC images. Thus, the cinema image and the NTSC image can be distinguished without any problem, and the coding can be improved.

As described above, the encoding side converts the NTSC frame rate into a cinema frame rate and performs encoding. On the decoding side, the data transmitted at the cinema frame rate is converted into the NTSC frame rate for displaying an image. Must be displayed. Therefore, information indicating a procedure for converting the cinema frame rate to the NTSC frame rate is added to the encoded data. In this case, when the data is converted and transmitted as shown in FIG. 3, information is added to the encoded data so as to perform the conversion from the cinema frame rate to the NTSC frame rate as shown in FIG.

That is, in FIG.
The frame and the second frame are respectively from the first and second frames of pack 4, the third frame of pack 5 is from the odd field of the second frame of pack 4, and the even field of the third frame of pack 4, and The fourth and fifth frames are respectively composed of the third frame and the fourth frame of the pack 4, and information for converting the cinema frame rate to the NTSC frame rate on the decoding side is added to the encoded data.

When converted and transmitted as shown in FIG. 4, a cinema frame rate as shown in FIG. 7 → NTSC
Information is added to the encoded data so as to convert the frame rate.

That is, in FIG.
The frames are from the first frame of pack 4, the second frame of pack 5 is the odd field of the first frame of pack 4, and the even field of the second frame of pack 4, the third, fourth and fifth frames of pack 5. Comprises the second, third and fourth frames of pack 4 respectively,
Information for converting the cinema frame rate to the NTSC frame rate on the decoding side is added to the encoded data.

Embodiment 2 FIG. FIG. 8 is a block diagram showing a digital video coding method according to the second embodiment of the present invention, and FIGS. 9 and 10 are diagrams showing resolution conversion. In the figure, reference numeral 105 denotes a post-conversion scene change detection unit, and 155 denotes a scene change detection signal. FIG. 1
The same reference numerals indicate the same or corresponding parts.

Next, the operation will be described with reference to FIGS. For example, 30 frames / s as input data 151
When the NTSC signal is input, the cinema detector 101 determines whether the input data 151 is originally a cinema sequence of 24 frames / s. The detected sequence is notified to the time resolution conversion unit 102 by the control signal 152. The sequence detected as a cinema sequence is subjected to time resolution conversion from 30 frames / s to 24 frames / s by predetermined processing by the time resolution conversion unit 102, and the result is transmitted as time resolution converted data 153. You. The transmitted time resolution converted data 153 is sent to the encoding unit 103 and the scene change detection unit 105. Scene change detection unit 105
Then, scene change detection is performed on the data 153 after the time resolution conversion, and a scene change detection signal 155 is sent to the encoding unit 103 for a portion determined to be a scene change. The encoding unit 103 converts the time resolution converted data 153 to 24 frames / s, and outputs a scene change detection signal 155 from the scene change detection unit 105.
For the sequence determined to be a scene change, the display order on the decoding side is included in the encoding information and transmitted as shown in FIG. 9 or FIG. For example, FIG.
In the case where the scene change occurs between the first frame and the second frame of the pack 4, the first and second frames of the pack 5 are respectively shifted from the first frame and the second frame of the pack 4 The three frames are decoded from the odd field of the second frame of Pack 4 and the even field of the third frame, and the fourth and fifth frames of Pack 5 are decoded and displayed from the third and fourth frames of Pack 4, respectively. Provide information to instruct the side.

In FIG. 10, when a scene change occurs between the second and third frames of pack 4,
The first frame of pack 5 is from the first frame of pack 4, and the second frame of pack 5 is from the odd field of the first frame of pack 4 and the even field of the second frame. The five frames each provide information instructing the decoding side to compose and display from the second, third, and fourth frames of pack 4.

Embodiment 3 FIG. 11 is a block diagram showing a digital video coding system according to the third embodiment of the present invention, and FIGS. 12 to 16 are diagrams showing resolution conversion. 104 is a pre-conversion scene change detection unit, 156, 1
57 is a scene change detection signal. The same reference numerals as those in FIG. 1 indicate the same or corresponding parts.

Next, the operation will be described with reference to FIGS. For example, 30 frames /
s NTSC signal is input, the input data 151
Originally, the cinema detection unit 101 determines whether the sequence is a cinema sequence of 24 frames / s. The detected sequence is transmitted to the time resolution conversion unit 102 by the control signal 152.
Is notified. On the other hand, the scene change detection unit 10
In 4, the scene change of the input data 151 is detected,
When a scene change occurs, the time resolution conversion unit 102
And the scene change detection signals 156 and 157 to the encoding unit 103, respectively. When the cinema sequence notification is obtained from the cinema detection unit 101 from the control signal 152, the time resolution conversion unit 102 normally obtains 30 frames / s → 24 as shown in the time resolution conversion on the encoding side in FIG.
Re-convert to frames / s. Further, when a scene change detection signal 156 is obtained from the scene change detection unit 104, when a scene change occurs between the first frame and the second frame of the pack 5 as shown in FIG. In this case, the first frame of the pack 5 constitutes the first frame of the pack 4, and the second, third, fourth and fifth frames of the pack 5 constitute the second, third and fourth frames of the pack 4. . By this method, even in the case where the time resolution conversion occurs near the scene change, the coding efficiency is improved because there is no frame in which the image before the scene change and the image after the scene change are mixed, and even in the vicinity of the scene change. High image quality can be obtained.

Similarly, when a scene change occurs between the second frame and the third frame of the pack 5 as shown in FIG. 12, the pack 4 is generated by dropping the second frame of the pack 5. By this method, even when the time resolution conversion occurs near the scene change,
Encoding efficiency is improved because there is no frame in which images before and after the scene change are mixed.
High image quality can be obtained even near a scene change.

Similarly, when a scene change occurs between the third frame and the fourth frame of pack 5 as shown in FIG. 13, the first frame of pack 4 becomes the first frame of pack 5.
From the frame, the second frame of pack 4 consists of the odd field of the third frame of pack 5 and the even field of the second frame, and the third and fourth frames of pack 4 consist of the fourth and fifth frames of pack 5, respectively. I do. By this method, even when the time resolution conversion occurs in the vicinity of the scene change, the coding efficiency is improved because there is no frame in which the images before and after the scene change are mixed. In this case, high image quality can be obtained.

When a sequence including a scene change is encoded by the above method, for example, when the time resolution conversion shown in FIG. 13 is performed on the encoding side, the display order on the decoding side is as shown in FIG. The first frame of the pack 5 is obtained from the first frame of the pack 4 and the second frame is obtained from the odd field of the first frame and the even field of the second frame of the pack 4, respectively. The frames are each composed of the second, third and fourth frames of pack 4. Thereby, 24 on the decoding side
When converting from frames / s to 30 frames / s, even in a scene change, there is no frame in which images before and after the scene change are mixed, so that a smooth video is subjectively obtained.

When a sequence including a scene change is encoded by the above method, for example, when the time resolution conversion shown in FIG. 12 is performed on the encoding side, the display order on the decoding side is as shown in FIG. The first and second frames of pack 5 are respectively separated from the first and second frames of pack 4,
The third frame of the pack 5 is obtained from the odd field of the second frame of the pack 4 and the even field of the third frame.
Of the third frame and the fourth frame. Thereby, when converting from 24 frames / s to 30 frames / s on the decoding side, since there is no frame in which the images before and after the scene change are mixed even in the scene change, a smooth video is subjectively displayed. can get.

[0042]

Since the present invention is configured as described above, it has the following effects.

Equipped with cinema detecting means and time resolution converting means, and by converting the time resolution from 30 frames / s to 24 frames / s, a cinema sequence can be efficiently encoded at the original 24 frames / s. And high-quality images can be obtained.

When converting the time resolution from 30 frames / s to 24 frames / s, if a cinema image and a normal NTSC (30 frames / s) image are mixed in the conversion target image, the conversion is performed efficiently. , Cinema and NTSC
A smooth image can be obtained even at the boundary of the image.

In the case of a cinema sequence, when converting from 30 frames / s to 24 frames / s, four frames (pack 4) are created in units of five frames (pack 5). Even when the cinema image is changed to the NTSC image signal in the second frame, the time resolution is converted so that there is no mixed frame of the cinema image and the NTSC image, so that a smooth image can be obtained even at the boundary between the cinema image and the NTSC image. Can be

In the case of a cinema sequence, when converting from 30 frames / s to 24 frames / s, four frames (pack 4) are created in units of five frames (pack 5). Even when the cinema image is changed to the NTSC image signal in the third frame, the time resolution conversion is performed so that there is no mixed frame of the cinema image and the NTSC image, so that a smooth image can be obtained even at the boundary between the cinema image and the NTSC image. Can be

In the case of a cinema sequence, when converting from 30 frames / s to 24 frames / s, four frames (pack 4) are created in units of five frames (pack 5). Even when the cinema image changes to an NTSC image signal in the fourth frame, a smooth image can be obtained even at the boundary between the cinema and the NTSC image because the time resolution is converted so that there is no mixed frame of the cinema image and the NTSC image. Can be

Further, the display order when the created 24 frames / s data is converted to a time resolution of 30 frames / s on the decoding side and displayed is written in the coded data, and further in the display order, the cinema image is used. Even when the signal is changed to an NTSC image signal, since the time resolution is converted so that a mixed frame of a cinema image and an NTSC image does not exist in displaying, a smooth image can be obtained even at the boundary between the cinema and the NTSC image.

A cinema detecting means and a time resolution converting means are provided. The time resolution is converted from 30 frames / s to 24 frames / s, and a scene change detection is performed on the data after the time resolution conversion by the scene change detecting means. By doing so, encoding can be efficiently performed at 24 frames / s at the time of a scene change, and a high-quality image can be obtained.

In the case of a cinema sequence, when converting from 30 frames / s to 24 frames / s, four frames (pack 4) are created in units of five frames (pack 5). Even when a scene change occurs in the cinema image in the fourth frame, the time resolution conversion is performed so that there is no mixed frame of the image before the scene change and the image after the scene change. An image is obtained.

Further, a cinema detecting means and a time resolution converting means are provided for performing a time resolution conversion from 30 frames / s to 24 frames / s. By performing detection, efficient 2
Encoding can be performed at 4 frames / s, and a high-quality video can be obtained.

When converting the time resolution from 30 frames / s to 24 frames / s, if there is a scene change in the image to be converted, the time resolution is converted so that there is no mixed frame before and after the scene change. Therefore, a smooth image can be obtained even in a scene change.

Further, the display order when the created 24 frames / s data is converted to a time resolution of 30 frames / s on the decoding side and displayed is written in the coded data, and further in the display order, the cinema image Even in the case of a scene change, a time-resolution conversion is performed so that a mixed frame before and after the scene change does not exist during display, so that a smooth image can be obtained even in the scene change.

[Brief description of the drawings]

FIG. 1 is a block diagram of a digital video coding system according to a first embodiment of the present invention.

FIG. 2 is a diagram showing an outline of time resolution conversion of a digital video coding method.

FIG. 3 is a diagram illustrating an example of a time resolution conversion method of a digital video coding method.

FIG. 4 is a diagram showing another example of the time resolution conversion method of the digital video coding method.

FIG. 5 is a diagram showing another different example of the time resolution conversion method of the digital video coding method.

FIG. 6 is a diagram illustrating an example of a time resolution conversion method on the decoding side of the digital video coding method.

FIG. 7 is a diagram illustrating another example of the time resolution conversion method on the decoding side of the digital video coding method.

FIG. 8 is a block diagram of a digital video coding system according to a second embodiment of the present invention.

FIG. 9 is a diagram illustrating an example of a time resolution conversion method on the decoding side of the digital video coding method.

FIG. 10 is a diagram illustrating another example of the time resolution conversion method on the decoding side of the digital video coding method.

FIG. 11 is a block diagram of a digital video coding system according to a third embodiment of the present invention.

FIG. 12 is a diagram illustrating an example of a time resolution conversion method of a digital video coding method.

FIG. 13 is a diagram illustrating another example of the time resolution conversion method of the digital video coding method.

FIG. 14 is a diagram illustrating an example of a time resolution conversion method on the decoding side of the digital video coding method.

FIG. 15 is a diagram illustrating another example of the time resolution conversion method on the decoding side of the digital video coding method.

FIG. 16 is a diagram showing another different example of the time resolution conversion method of the digital video coding method.

FIG. 17 is a diagram showing a time resolution conversion method of a conventional digital video coding method.

[Explanation of symbols]

101 cinema detector, 102 time resolution converter, 1
03 encoding unit, 104 scene change detection unit before conversion, 105 scene change detection unit after conversion, 151 input image data, 152 time resolution conversion instruction control signal,
153 Data after time resolution conversion, 154 encoded data, 155 scene change detection signal.

Claims (16)

[Claims] 1. A digital video coding system for coding a video signal, comprising: a cinema detection means for determining whether input data is a normal NTSC sequence or a cinema sequence created at a cinema frame rate; Removes redundant fields from the sequence determined as a cinema sequence by
Time resolution conversion that performs time resolution conversion from the TSC frame rate to the cinema frame rate, and performs time resolution conversion that optimizes coding efficiency at the boundary between a normal NTSC sequence and a cinema sequence or at a scene change portion in a cinema sequence. Means and instruction information for performing a time resolution conversion on the decoding side from the cinema frame rate to the NTSC frame rate for displaying at an NTSC frame rate when decoding a cinema sequence converted to a cinema frame rate at the time of decoding. Encoding means for adding a character string to encoded data. 2. The time resolution conversion means performs time conversion on input data of an NTSC frame rate in a specific unit (hereinafter referred to as pack 5) in encoding-side resolution conversion (unit after pack 5 is time-converted). Is hereinafter referred to as pack 4),
Cinema sequence and normal NTS in pack 5
2. The digital moving picture coding method according to claim 1, wherein when the sequence of C is included, the conversion from pack 5 to pack 4 is performed by performing a specific process such that the coding efficiency is not deteriorated. 3. When the pack 5 is composed of 5 frames and the pack 4 is composed of 4 frames, the time resolution conversion means determines that the first frame in the pack 5 is a cinema sequence, and , When the fourth and fifth frames are the normal NTSC sequence, the first frame of pack 5 to the first frame of pack 4, the second frame of pack 5 to the second frame of pack 4, and the fourth frame of pack 5 3
3. The digital moving image according to claim 2, wherein a third frame of the pack 4 is formed from an even field of the frame and an odd field of the fourth frame, and a fourth frame of the pack 4 is formed from the fifth frame of the pack 5. Image coding method. 4. The time resolution conversion means, when the pack 5 is composed of 5 frames and the pack 4 is composed of 4 frames, the first and second frames in the pack 5 are cinema sequences and the third 3. The digital moving picture coding method according to claim 2, wherein when the fourth and fifth frames are normal NTSC sequences, the second frame of the pack 5 is deleted to form the pack 4. 5. The time resolution conversion means, wherein when the pack 5 is composed of 5 frames and the pack 4 is composed of 4 frames, the first, second and third frames in the pack 5 are cinema sequences, When the fourth and fifth frames are normal NTSC sequences, the first, second, and third frames of pack 5 are converted to the first and second frames of pack 4, and the fourth and fifth frames of pack 5 are converted. 3. The digital moving picture coding method according to claim 2, wherein the conversion into the third and fourth frames of the pack 4 is performed. 6. The encoding means, when decoding encoded cinema frame rate data, when converting the cinema frame rate to the NTSC frame rate on the decoding side, the conversion from pack 4 to pack 5 5 are converted from the first frame of pack 4 and the second frame of pack 4, respectively,
Are created from the odd field of the second frame of pack 4 and the even field of the third frame of pack 4, and the fourth and fifth frames of pack 5 are obtained from the third and fourth frames of pack 4, respectively. 3. The digital moving picture coding method according to claim 2, wherein information to be created is added to the coded data. 7. When the encoding means decodes encoded data at a cinema frame rate, when the decoding side converts the cinema frame rate to the NTSC frame rate, the conversion from pack 4 to pack 5 is performed at pack time. The first frame of pack 5 is converted to the first frame of pack 4, the second frame of pack 5 is created from the odd field of the first frame of pack 4 and the even field of the second frame of pack 4, 3. The digital dynamic encoding system according to claim 2, wherein the third, fourth, and fifth frames add information created from the second, third, and fourth frames of the pack 4 to the encoded data. 8. In a digital moving picture coding method for coding a moving picture signal, a cinema detecting means for determining whether input data is a normal NTSC sequence or a cinema sequence created at a cinema frame rate, and the cinema detecting means Removes redundant fields from the sequence determined as a cinema sequence by
Encoding-side time resolution conversion means for performing a time resolution conversion from a TSC frame rate to a cinema frame rate; converted scene change detection means for performing a scene change detection on the sequence after the time resolution conversion; A digital video encoding system comprising: encoding means for adding, to encoded data, temporal resolution conversion instruction information for displaying an image decoded at a temporal resolution of a cinema sequence in a normal NTSC sequence. 9. The encoding means, when the scene change detection means after conversion detects that a scene change has occurred between the first frame and the second frame in the pack 4, performs decoding when decoding. 4 frames from pack 4 to NT
When converting the time resolution to pack 5 of 5 frames of SC,
The first frame of pack 5 is from the first frame of pack 4, the second frame of pack 5 is from the second frame of pack 4, and the third frame of pack 5 is the odd field and third frame of the second frame of pack 4. The fourth frame of the pack 5 is added with the instruction information to the encoded data so as to be created from the third frame of the pack 4, and the fifth frame of the pack 5 is created from the fourth frame of the pack 4. The digital moving picture coding method according to claim 8, wherein 10. The encoding means, when the scene change detection means after conversion detects that a scene change has occurred between the second frame and the third frame in the pack 4,
At the time of decoding, packs 4 to N
When converting the time resolution to pack 5 of 5 frames of TSC, the first frame of pack 5 is from the first frame of pack 4, and the second frame of pack 5 is the odd field of the first frame of pack 4 and the second frame of pack 2. From Evenfield, the third frame of pack 5 is from the second frame of pack 4, and the fourth frame of pack 5 is the third frame of pack 4.
9. The digital moving picture encoding method according to claim 8, wherein the instruction information is added to the encoded data so that the fifth frame of the pack 5 is created from the fourth frame of the pack 4 from the frame. 11. In a digital moving picture coding method for coding a moving picture signal, input data is a normal NTSC signal.
A cinema detecting means for determining whether the sequence is a cinema sequence created at a cinema frame rate, and removing a redundant field from the sequence determined to be a cinema sequence by the cinema detecting means, and converting a cinema frame from the NTSC frame rate of the input data. Encoding-side time resolution conversion means for performing time resolution conversion to a rate,
A pre-conversion scene change detecting means for detecting a scene change with respect to a sequence before being subjected to time resolution conversion;
Encoding means for adding an instruction of time resolution conversion to be displayed at a frame rate to encoded data. 12. The encoding-side time resolution conversion means includes:
When the SC frame rate is 30 frames / s and the cinema frame rate is 24 frames / s, pack 5 is composed of 5 frames, pack 4 is composed of 4 frames, and the first frame in pack 5 is When a scene change is detected between the second frames, when converting from pack 5 to pack 4, the first frame of pack 5 to the first frame of pack 4 are formed, and the second, third, From the fourth and fifth frames, the second and third packs 4
12. The digital moving picture coding method according to claim 11, wherein the digital moving picture coding method comprises a fourth frame and a fourth frame. 13. The encoding-side time resolution conversion means, wherein the pack 5 is composed of 5 frames, the pack 4 is composed of 4 frames, and a scene change between the second frame and the third frame in the pack 5 is performed. 12. The digital moving picture coding method according to claim 11, wherein, when detected, when converting from pack 5 to pack 4, the second frame of pack 5 is deleted to form pack 4. 14. The encoding-side temporal resolution conversion means is arranged such that the pack 5 is composed of 5 frames, the pack 4 is composed of 4 frames, and a scene change between the third frame and the fourth frame in the pack 5 is performed. If detected, when converting from pack 5 to pack 4, the first, second, and third frames of pack 5 are converted to the first and second frames of pack 4, and the fourth and fifth frames of pack 5 are converted. 12. The digital moving picture coding method according to claim 11, wherein the digital moving picture is converted into the third and fourth frames of pack 4. 15. When encoding data of a cinema frame rate is decoded, the encoding means converts the pack 4 to pack 5 when converting the cinema frame rate to the NTSC frame rate on the decoding side. First of 5
And the second frame are converted from the first and second frames of pack 4, the third frame of pack 5 is created from the odd field of the second frame of pack 4, and the even field of the third frame of pack 4, and 12. The digital moving picture coding method according to claim 11, wherein instruction information is added to the coded data so as to create the fourth and fifth frames of the pack 4 from the third and fourth frames. 16. The encoding means, when decoding encoded cinema frame rate data, performs conversion from pack 4 to pack 5 when converting the cinema frame rate to the NTSC frame rate on the decoding side. First of 5
The frame is converted from the first frame of pack 4, the second frame of pack 5 is created from the odd field of the first frame of pack 4, and the even field of the second frame of pack 4, and the third and fourth frames of pack 5 are formed. 12. The digital moving picture coding method according to claim 11, wherein instruction information is added to the coded data so that the fifth and fifth frames are created from the second, third, and fourth frames of the pack 4.