Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
1998-12-23
2002-10-15
Rao, Andy (Department: 2613)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
C375S240270, C375S240250
Reexamination Certificate
active
06466625
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a picture signal processing method and apparatus used with advantage for recording/reproducing moving picture signals and still picture signals on or from a recording medium, such as an optical disc or a magnetic tape, and for transmitting moving and still picture signals from a sender to a receiver over a transmission path such as in a teleconferencing system, a television telephone system or in a broadcasting equipment.
2. Description of the Related Art
As an encoding system for efficient encoding of still or moving pictures, by way of data compression, there has so far been known a block encoding system, such as block discrete cosine transform (DCT) encoding system.
In companding picture data by encoding, the ringing noise (mosquito noise) or block distortion tends to be produced. The higher the compression ratio, the more readily the noise is liable to be produced. Also, the more intricate the input picture data, the more readily the noise is liable to be produced. The result is that deterioration due to the noise is produced in a picture corresponding to the picture data obtained on compression/expansion.
For reducing picture deterioration of the picture data due to noise produced on compression/expansion, it may be contemplated to use a pre-filter, such as a low-pass filter (LPF), and to perform filtering on the picture data corresponding to the input picture prior to encoding.
FIG. 1
shows an example of an encoding device for moving picture signals using the above-mentioned pre-filter.
In the encoding device, shown in
FIG. 1
, digital moving picture signals, entering a terminal
201
, are filtered by a pre-filtering circuit
202
. The filtered picture signals for processing are entered to an encoding circuit
205
.
The encoding circuit
205
encodes digital moving picture signals using a hybrid encoding method combined from the motion-compensated inter-frame prediction and transform coding, such as DCT. That is, the encoding circuit
205
performs inter-frame/intra-frame adaptive prediction on an input signal S
211
, while performing DCT on prediction error signals by an arithmetic unit
206
. A quantizer
208
quantizes calculated DCT coefficients to output an encoded output signal S
215
at a terminal
220
. It is noted that the quantizer
208
sets the quantization step size which will give a constant bitrate of the output signal S
215
.
The output signal S
215
is dequantized by a dequantizer
209
and inverse DCTed by an inverse DCT unit
210
to restore prediction error signals to which prediction picture signals are added at an arithmetic unit
211
by way of local decoding. This addition gives decoded picture signals S
213
which are stored in a frame memory
212
.
In general, the more intricate the input moving picture or the more vigorous its motion, the larger becomes an inter-frame difference signal S
212
. If, for suppressing the volume of the generated information to a pre-set bitrate, coarse quantization is used at this time at the quantizer
208
, perceptually outstanding block distortion is produced to deteriorate the apparent picture quality.
Thus, in the encoding device shown in
FIG. 1
, the low-pass passband of the pre-filter with respect to the input picture is restricted, in dependence upon the magnitude of the inter-frame difference signal of the input picture, responsive to the magnitude of the inter-frame difference signal of an input picture, in consideration that the magnitude of such inter-frame difference signal of the input signal affects the quantity of generated bits for inter-frame predictive coding. This attenuates the energy of the prediction error signals to prohibit coarse quantization to improve the apparent picture quality.
The variable pre-filter controlling method in the present encoding device is now explained.
In the arithmetic unit
204
, an inter-frame difference r between a picture signal S
210
entering the input terminal
201
and a picture signal S
214
entering a memory
212
is calculated. The picture signal S
214
is a signal to which reference is had in the inter-frame prediction for the input picture signal S
210
. The inter-frame difference r is entered to a pre-filter controller
203
which then generates a pre-filtering coefficient k, as a parameter for controlling the low-pass passband of the pre-filter, in dependence upon the magnitude of the inter-frame difference r. The pre-filter
202
has two-dimensional low-pass filter characteristics, shown in
FIG. 2
, with the low-pass passband showing monotonous increase with respect to the pre-filtering coefficient k.
FIG. 3
shows the relation between the inter-frame difference r and the pre-filtering coefficient k. By the above relation, the pre-filter controller
203
controls the low-pass characteristics of the pre-filter depending on the magnitude of the inter-frame difference signal of the input picture.
As a technique for reducing the mosquito noise and the block distortion following the decoding, there is known a technique employing a system in which the pixel-based noise quantity is predicted from the macroscopic characteristics of the block level and from the microscopic characteristics of the pixel level, based on the local statistic quantity and the encoding information of the picture, and the noise component is removed by adaptive filtering.
As the technique for reducing the block distortion or the mosquito noise, there is known a post-filtering technique of filtering as-decoded output picture signals in addition to the above-described technique.
Although the above-mentioned pre-filtering processing, prior to encoding, is able to suppress the mosquito noise, the block distortion cannot be effectively removed especially in a picture performing vigorous movements. Moreover, the high frequency components are diminished by pre-filtering, such that, if a pre-filtered picture is decoded, only the block distortion is apparent to the eye.
In the post-filtering processing, performed following the decoding, there is no change caused in the encoding efficiency following the encoding, because no processing occurs on the decoding side, such that it is not possible to cope successfully with roughed quantization over an entire input picture performing vigorous movements difficult to predict.
That is, in a conventional picture encoding/decoding apparatus, the pre-filtering prior to the encoding and the post-filtering following the decoding are performed independently of each other, such that the noise cannot be removed effectively.
SUMMARY OF THE INVENTION
It is therefore an object of the present invention to provide a picture signal processing method and apparatus whereby it is possible to suppress the block distortion or the mosquito noise generated on encoding or decoding input moving or still pictures.
In one aspect, the present invention provides a picture signal processing apparatus for encoding and decoding an input picture on a block basis, including filtering means for filtering the input picture responsive to characteristics of the input picture, encoding means for block encoding a picture filtered by the filtering means, decoding means for decoding the picture encoded by the encoding means and noise reducing means for reducing at least the block distortion of the picture decoded by the decoding means.
With this picture signal processing apparatus, the input is filtered by a filtering means before being passed through the encoding means, while the picture decided by the decoding means is reduced in block distortion by noise reducing means.
With the picture signal processing apparatus, the input is improved in encoding efficiency by filtering, while it is reduced in the block distortion or mosquito noise otherwise produced on decoding.
Thus, with the present picture signal processing apparatus, the high frequency components of the input picture are reduced and the block distortion etc produced on decoding is rendered apparent to improve the detection accuracy of the block distort
Igi Nobuhiro
Kato Motoki
Kobayashi Hiroshi
Morio Minoru
Obata Koji
Frommer William S.
Frommer & Lawrence & Haug LLP
Rao Andy
Savit Glenn F.
LandOfFree
Picture signal processing method and apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Picture signal processing method and apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Picture signal processing method and apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2975900