JPH0479586A - Picture quality deterioration detection system for television signal - Google Patents
Picture quality deterioration detection system for television signalInfo
- Publication number
- JPH0479586A JPH0479586A JP2190505A JP19050590A JPH0479586A JP H0479586 A JPH0479586 A JP H0479586A JP 2190505 A JP2190505 A JP 2190505A JP 19050590 A JP19050590 A JP 19050590A JP H0479586 A JPH0479586 A JP H0479586A
- Authority
- JP
- Japan
- Prior art keywords
- pixel
- error
- minimum value
- pixels
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 230000006866 deterioration Effects 0.000 title claims description 10
- 238000001514 detection method Methods 0.000 title description 12
- 230000005540 biological transmission Effects 0.000 claims abstract description 29
- 238000012937 correction Methods 0.000 claims abstract description 17
- 238000000034 method Methods 0.000 claims description 47
- 238000013139 quantization Methods 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 abstract description 3
- 230000000052 comparative effect Effects 0.000 abstract 2
- 230000003044 adaptive effect Effects 0.000 description 6
- 230000000694 effects Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 241000677647 Proba Species 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Landscapes
- Detection And Correction Of Errors (AREA)
- Color Television Systems (AREA)
- Television Systems (AREA)
- Facsimile Image Signal Circuits (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
- Error Detection And Correction (AREA)
Abstract
Description
【発明の詳細な説明】
(発明の技術分野)
本発明は、動画像符号化・復号化方式における、特に高
速符号化伝送を必要とする高精細度テレビジョン(以降
、HDTVという)等の伝送路誤りによる画質劣化検出
方式に関する。Detailed Description of the Invention (Technical Field of the Invention) The present invention relates to the transmission of high-definition television (hereinafter referred to as HDTV), etc., which particularly requires high-speed encoding and transmission, in a moving image encoding/decoding system. This invention relates to a method for detecting image quality deterioration due to path errors.
(従来の技術)
動画像信号の符号化伝送において、伝送路誤りによって
復号側で検知するスパイク雑音等は、受信側て誤つ画素
をピークとする画質劣化を生ずる。(Prior Art) In encoded transmission of a moving image signal, spike noises detected on the decoding side due to transmission path errors cause image quality deterioration on the receiving side, with the peak occurring at the erroneous pixel.
例えば、その伝搬雑音はフィールド内1次元予測のとき
第1図(b)の例のごとく水平方向への1次元雑音とし
て、2次元予測のときは斜め右下方への2次元雑音して
、画質劣化を生じさせる。For example, in the case of intra-field one-dimensional prediction, the propagation noise is one-dimensional noise in the horizontal direction as shown in the example in Figure 1(b), and in the case of two-dimensional prediction, it is two-dimensional noise diagonally downward to the right, resulting in image quality. cause deterioration.
この画質劣化対策としては、■誤り画素を付加情報とし
て受信した誤り訂正符号から計算再生した正常画素と置
換する誤り訂正符号方式と、■付加情報なくして誤り訂
正符号と独立に誤りを検出するため画素に隣接した正常
画素を参照して画素値を見なし復元するポストフィルタ
処理構成をとる画質劣化検出法(エラーコンシールメン
ト法、以下EC法と呼ぶ)かある。Measures against this image quality deterioration include: ■ an error correction code method in which erroneous pixels are replaced with normal pixels calculated and reproduced from an error correction code received as additional information; and ■ a method to detect errors independently of the error correction code without additional information. There is an image quality deterioration detection method (error concealment method, hereinafter referred to as EC method) that takes a post-filter processing configuration in which pixel values are considered and restored by referring to normal pixels adjacent to the pixel.
(発明が解決しようどする課題)
前者は正確に誤り訂正が行える反面、誤り訂正符号の様
に付加的な伝送情報量を必要とする。後者は前者と比較
して誤り画素位置検出の精度かあまり高くなく画質の改
善と劣化を同時にもたらすという難点がある反面、一般
的に付加情報量なしで誤り位置を推定する。(Problems to be Solved by the Invention) Although the former method allows accurate error correction, it requires an additional amount of transmission information like an error correction code. The latter method has the disadvantage that the accuracy of detecting the position of an error pixel is not so high as compared to the former method, and it simultaneously improves and degrades the image quality. However, on the other hand, the error position is generally estimated without any amount of additional information.
現在、画像符号化・復号化装置は、伝送路誤りに対する
対策として両方式の併用が行なわれるのが一般的である
。Currently, image encoding/decoding devices generally use both methods in combination as a countermeasure against transmission path errors.
情報量の削減を行なっていても大容量の伝送量を必要と
するH D T Vの伝送方式ては、誤り訂正符号を多
用すると伝送効率が悪くなるためE、 C法は画質向上
のため必須てあり、EC法による伝送路誤りての画質劣
化の判定精度を高めることが重要である。In the HDTV transmission system, which requires a large amount of data to be transmitted even if the amount of information is reduced, the use of error correction codes reduces transmission efficiency, so methods E and C are essential for improving image quality. Therefore, it is important to improve the accuracy of determining image quality degradation due to transmission path errors using the EC method.
しかしながら、従来のEC法、例えばスパイク雑音除去
に有効な単純なメデイアンフィルタ方式%式%
35、pp646−652. (Jun−1987)
) 、またはスパイク雑音におけるエツジ成分の大小比
較方式では、判定精度か低くかつ正常画素を誤り画素と
誤判定する確率も高いため、本質的な画質向上にならな
い。However, conventional EC methods, for example, a simple median filter method effective for removing spike noise, are not effective.35, pp646-652. (Jun-1987)
), or a method of comparing the magnitude of edge components in spike noise, the determination accuracy is low and the probability of erroneously determining a normal pixel as an error pixel is high, so it does not substantially improve image quality.
すなわち、従来のEC法における単純なメデイアンフィ
ルタ法ては、エツジ成分の大きさに関わらず、その大き
さを画素群の中間値で置き返ることにより抑圧するフィ
ルタ処理どなる。これにより、エツジ成分とならない正
常画素もこのフィルタ処理により誤りと判定され”〔し
よう。また、発明の効果で触れている方式1(メティア
ンフィルタを利用)でもエツジ部の大きさ判定を行わな
いため、誤判定の確率か高い。That is, the simple median filter method in the conventional EC method is a filtering process that suppresses the edge component by replacing its size with an intermediate value of a group of pixels, regardless of the size of the edge component. As a result, normal pixels that do not become edge components are also determined to be errors through this filtering process.Furthermore, even in method 1 (using a metian filter) mentioned in the Effects of the Invention, the size of the edge portion is not determined. Therefore, the probability of misjudgment is high.
また、従来のEC法であるエツジ部の大小判定では、そ
のエツジ部の符号判定か行われないため、誤りが画素が
例えば3画素中の中間値てあってもピークの画素として
誤判定してしまう。In addition, in the conventional EC method, which determines the size of an edge part, only the sign of the edge part is determined, so even if a pixel is an intermediate value among three pixels, it may be incorrectly determined as a peak pixel. Put it away.
本発明は、上述の難点を克服するためのものて、誤り訂
正符号のような付加情報増加なせずにかつ誤り画素の検
出精度を富め、復号化画像の品質向上を容易に達成てき
る画質劣化の検出方式を提供することを目的とする。In order to overcome the above-mentioned difficulties, the present invention improves the accuracy of detecting error pixels without increasing additional information such as an error correction code, and easily improves the quality of decoded images. The purpose of this paper is to provide a method for detecting deterioration.
(課題を解決するための手段)
本発明の特徴は、伝送路誤りによるスパイク雑音をもつ
画素(以降、誤り画素という)と誤り画素に隣接した正
常画素との間に、エツジ成分がある確率をもって一定の
振幅で存在することに着目17、エツジ成分の存在を検
出するエツジ成分(正負符号)判定と、量子化特性か有
するエラー伝搬振幅で決定される閾値との大小比較によ
り伝送路誤りを推定する検出方式である。(Means for Solving the Problems) A feature of the present invention is that there is a probability that an edge component exists between a pixel having spike noise due to a transmission path error (hereinafter referred to as an error pixel) and a normal pixel adjacent to the error pixel. Focusing on the fact that the edge component exists with a constant amplitude17, the transmission path error is estimated by comparing the magnitude of the edge component (positive/negative sign) judgment to detect the presence of the edge component with a threshold determined by the error propagation amplitude that has the quantization characteristic. This is a detection method that
本発明は、簡易なアルゴリスムによりハードウェアの小
型化を促すとともに、)−I D T Vなどの高速な
画像処理に対応しやすい特徴をも持つ。The present invention promotes miniaturization of hardware by using a simple algorithm, and also has the feature of being easily compatible with high-speed image processing such as )-ID TV.
(原理)
第1図は、本方式の説明を行うものである。図[ト弘方
式説明の簡単化のため、対象画素として3画素に限定さ
れた隣接画素a、 b、 cを同一フィールド」二の画
素とし、Xのみが伝送路誤りによってスパイク雑音を持
つ誤り画素である。隣接画素として同一・フィール1へ
中において水平、垂直画素に加え、斜め方向の画素も参
照するが本説明では水平、垂直画素て説明する。(Principle) FIG. 1 explains this method. Figure [To simplify the explanation of the Tohiro method, the target pixels are limited to 3 adjacent pixels a, b, and c, which are in the same field. Only X is an error pixel that has spike noise due to transmission path errors. It is. In addition to horizontal and vertical pixels within the same field 1 as adjacent pixels, diagonal pixels are also referred to, but in this description, horizontal and vertical pixels will be explained.
一次元フイールド内予測方式て考察すると、第1図(b
)のごとく誤り画素Xは、符号化方向である右方向に誤
り画素Xをピークにスパイク的な雑音として尾を引くこ
とになり、次に復号化される画質を劣化させ、さらその
次に符号化されるべく画素も順次に劣化させる。Considering the one-dimensional intra-field prediction method, Figure 1 (b
), the error pixel X leaves a trail as a spike-like noise in the right direction, which is the encoding direction, with the peak at the error pixel The pixels are also sequentially degraded in order to improve the image quality.
Xを誤り画素か否か判定するには、次式(1)のように
Xの隣接画素a、、b、cを用い、まずエツジ成分(X
−a) 、 (X−b) 、 (X−c)の絶対値の最
小値がある閾値より大きく、さらにこれら成分の正負符
号がすべて同一のとき、Xを誤り画素と判定する。To determine whether or not X is an error pixel, first use the edge component (X
-a) When the minimum absolute value of (X-b) and (X-c) is larger than a certain threshold value, and furthermore, the positive and negative signs of these components are all the same, X is determined to be an error pixel.
m1n(l x−a l 、 1x−b l 、l
x−cl ) >Th (1)(ただし、 (x−
a)、(x−b)、(x−c)<OまたLt (x−
a)、(x−b)。m1n(l x-a l, 1x-b l, l
x-cl ) > Th (1) (where (x-
a), (x-b), (x-c)<O or Lt (x-
a), (x-b).
(x−c)>0)
閾値Thは各量子化特性が有するエラー伝搬振幅より決
定される。(x-c)>0) The threshold Th is determined from the error propagation amplitude of each quantization characteristic.
すなわち、具体的には雑音となる正常画素のエツジ成分
の大きさを判定するものてあり、あらかしめ量子化特性
に基づく伝搬エラーによる雑音エッジ部の高さをシミュ
レーションで求めて置くことて決める。Specifically, it is used to determine the magnitude of the edge component of a normal pixel that becomes noise, and is determined by calculating and setting the height of a noise edge portion due to a propagation error based on preliminary quantization characteristics through simulation.
誤り画素と判定した画素は、隣接画素を用いて従来技術
にある内挿補間法等により画素Xを補間する。For the pixel determined to be an error pixel, pixel X is interpolated using adjacent pixels by an interpolation method or the like in the prior art.
なお、隣接画素として説明上用いた画素は第1図(a)
で見る直上(a)、左隣(b)、直下(c)を用いたか
、右」二、直上、左上、左隣、布下、直下、左下のいず
れの画素を絹み合わゼてもよいことばいうまてもない。The pixels used for explanation as adjacent pixels are shown in FIG. 1(a).
Words that can be used to combine any of the pixels directly above (a), next to the left (b), and below (c) as seen in Needless to say.
(実施例) 第2図は、本発明を適用した装置のソロツク図である。(Example) FIG. 2 is a solo diagram of an apparatus to which the present invention is applied.
伝送路を通過してきた符号化信号は復号部1で復号化さ
れ、再生画素としてフレームメモリに逐次蓄積する。フ
レームメモリの容量は第1図(a)で示ず画素Cから画
素aまでの画素を最低限蓄積出来れは゛よい。The encoded signal that has passed through the transmission path is decoded by a decoding unit 1 and sequentially stored in a frame memory as reproduced pixels. The capacity of the frame memory is not shown in FIG. 1(a), but it is best if it can store at least the pixels from pixel C to pixel a.
2ライン遅延部2ては画素aをフレームメモリ15から
読み込み、1ライン遅延部3は画素すをフレームメモリ
15から読み込み、1画素遅延部4は画素aを1ライン
遅延部3から受は取り、画素読み込み部16は画素Cを
フレームメモリ15から読み込む。The 2-line delay unit 2 reads the pixel a from the frame memory 15, the 1-line delay unit 3 reads the pixel a from the frame memory 15, the 1-pixel delay unit 4 receives and receives the pixel a from the 1-line delay unit 3, The pixel reading unit 16 reads the pixel C from the frame memory 15.
読み込まれた画素a、 b、 c、 Xを基に減算器5
.6.7ては(x−a) 、 (y、−b) 、 (x
−c)をそれぞれ計算し、その結果を絶対値計算部9、
l0111およびsign判定部判定用8する。sig
n判定部判定用8y、−a)、(x−b)、(x−c)
<OまたU (x−a)、(x−b)、(x−c)>
0の条件をチエツクし、その結果を最小値選択部12に
出力する。Subtractor 5 based on the read pixels a, b, c, and
.. 6.7 So (x-a), (y,-b), (x
-c) respectively, and the results are sent to the absolute value calculation unit 9,
l0111 and sign determination section 8 for determination. sig
n judgment unit judgment 8y, -a), (x-b), (x-c)
<O or U (x-a), (x-b), (x-c)>
The condition of 0 is checked and the result is output to the minimum value selection section 12.
最小値選択部12では、sign判定部判定用8により
最小値を比較判定部13に出力し、比較判定部13では
最小値選択部から最小値と閾値Thとを比較、条件満足
ならば誤り補正部14に対して画素Xの補正指示信号を
出力する。In the minimum value selection section 12, the sign judgment section judgment 8 outputs the minimum value to the comparison judgment section 13, and the comparison judgment section 13 compares the minimum value from the minimum value selection section with the threshold Th, and if the condition is satisfied, error correction is performed. A correction instruction signal for pixel X is output to section 14.
誤り補正部14では、補正指示信号を受けると従来技術
である内挿補間法により画素Xの補正を行う。Upon receiving the correction instruction signal, the error correction unit 14 corrects the pixel X using the conventional interpolation method.
(発明の効果)
本方式の効果を、伝送路誤りによる劣化が特定長て収束
する伝搬特性を有している予測値適応型量子化の伝送路
誤り伝搬特性について応用した場合について説明する。(Effects of the Invention) A case will be described in which the effects of this method are applied to the transmission path error propagation characteristics of predictive value adaptive quantization, which has a propagation characteristic in which deterioration due to transmission path errors converges over a specific length.
なお、本方式は予測適応型量子化の誤り伝搬防止に特に
有効である。Note that this method is particularly effective in preventing error propagation in predictive adaptive quantization.
予測値適応型量子化は、各予測値に基づいて適応的に量
子化特性が変更てきる特長を有するため、MaxやRe
flected 量子化に比べよりきめ細かな量子化特
性が実現できる。Predicted value adaptive quantization has the feature that the quantization characteristics can be adaptively changed based on each predicted value, so Max and Re
More fine-grained quantization characteristics can be achieved compared to reflected quantization.
さらに、この量子化は予測値に対し予測誤差を量子化す
るのでなく、符号化されるべき入力画素値そのものを量
子化し符号化割当することが可能となる特長を有する。Furthermore, this quantization has a feature that it is possible to quantize the input pixel value itself to be encoded and assign it to encoding, instead of quantizing the prediction error with respect to the predicted value.
これにより、通常のDPCMで問題となっている伝送路
誤りの伝搬が大幅に改善される。すなわち、伝送路誤り
が生しても次の送信コートが正しく受信されれば、正常
な予測値に依存した再生値が作成されるため、予測誤差
画素列に容易に戻すことか可能どなる。伝送路誤りは、
復号化側の再生値が変わったことによる量子化特性の差
たけてあり、その伝搬長は極めて短いといえる。This greatly improves the propagation of transmission path errors, which is a problem in normal DPCM. That is, even if a transmission path error occurs, if the next transmission code is correctly received, a reproduced value that depends on a normal predicted value is created, so it is not possible to easily restore the predicted error pixel string. Transmission path error is
The difference in quantization characteristics is due to the change in the reproduction value on the decoding side, and the propagation length can be said to be extremely short.
この予測値適応量子化を用いて実画像によるシミュレー
ション実験により効果を示す。実験の予測方式は1次元
フィールド内前値予測、量子化器は4ビット/画素のY
信号用予測値適応型量子化とした。簡略化のため伝送路
誤りは、ランダムビットエラーとし、さらに4ビツトコ
ードのうち同時に2ビット以上の誤りは発生しないと仮
定した。Using this predictive value adaptive quantization, we demonstrate its effectiveness through simulation experiments using real images. The prediction method used in the experiment was one-dimensional intra-field prior value prediction, and the quantizer was Y of 4 bits/pixel.
We used predictive value adaptive quantization for signals. For simplification, it is assumed that transmission path errors are random bit errors, and that errors in two or more bits of a four-bit code do not occur simultaneously.
実験には、3種のHDTVテスト動画像(flow−e
r、room float)を用いた。表1に、各ビッ
ト(MSB・・・LSB)に発生させた伝送路誤りに対
する誤り振幅の伝搬特性を示す。この結果及び画質評価
より、第2 MSB以後の誤りは殆ど伝搬しないことが
明かとなったため、第1 MSBての伝送誤り対策だけ
を考えることにする。また、復号化側にて、ビット誤り
による誤り量がピークとなる画素が検出されれば、一定
長の短い誤り伝搬長どなるのて、一定長画素列のみに対
して補正処理をすればよいことになる。In the experiment, three types of HDTV test video images (flow-e
r, room float) was used. Table 1 shows the propagation characteristics of error amplitudes for transmission path errors generated in each bit (MSB...LSB). From this result and image quality evaluation, it has become clear that errors after the second MSB are hardly propagated, so we will only consider countermeasures against transmission errors for the first MSB. Additionally, if the decoding side detects a pixel where the amount of error due to bit errors peaks, it is only necessary to perform correction processing on a constant length pixel string, regardless of the short error propagation length. become.
表コニ伝送路誤りに対する誤り振幅の伝搬特性4 ビッ
ト/ビクセルコード (X 1/255)方式2は
、本発明での方式であり、人間の視覚特性の観点から特
に平坦部での大振幅雑音の検出確率向上を目的としたも
のである。Propagation Characteristics of Error Amplitude for Transmission Path Errors 4 Bit/Vixel Code (X 1/255) Method 2 is the method used in the present invention, and from the viewpoint of human visual characteristics, it is particularly effective for large amplitude noise in flat areas. The purpose is to improve detection probability.
表2に、これらによる第1の誤り画素の正検出確率と正
常画素の誤検出確率の結果を示す。予測値適応量子化に
対する場合、方式2(本発明)か方式1に比べ正検出確
率て約10%高く、誤検出ては逆に約1/10に減少す
る優れた特性となった。Table 2 shows the results of the correct detection probability of the first error pixel and the false detection probability of the normal pixel. In the case of predicted value adaptive quantization, the probability of correct detection is about 10% higher than that of method 2 (the present invention) or method 1, and the probability of false detection is reduced to about 1/10, which is an excellent characteristic.
表2 検出確率
そこで、伝送路誤りにおける振幅のピークとなる第1番
目の誤り画素を検出するため、簡易かつ特性の優れた以
下の2方式を示す。Table 2 Detection Probability Therefore, in order to detect the first error pixel that is the peak of the amplitude in a transmission path error, the following two methods are shown which are simple and have excellent characteristics.
方式1ばインパルス雑音検出に優れたものとして報告さ
れている従来方式メデイアンフィルタ法(次式参黒)で
ある。Method 1 is the conventional median filter method (see the following formula), which has been reported to be excellent in detecting impulse noise.
x−Meclian(a、x、c) 1> a l
max(a、x、c)−min(a、x、c) l
(0<α<1)a、b、c:xの隣接画素(xi 、
j−1,xi−1゜JIX11J+1) )
Pcd:C:orrect detect proba
bility of 1−st errorPmd:M
iss detect probability o
f no−error方式1については、誤検出確率に
よる画質劣化が多いため全画素に対し判定し、注目画素
と」1下の画素を含む3画素を用いたMedianフィ
ルターにて内挿補間を行った。一方、方式2(本発明)
については、正検出確率が高くかつ誤検出が低いため、
誤りと判定された画素を含む約15画素に対してのみ方
式1と同じ内挿補間した。x-Meclian (a, x, c) 1> a l
max (a, x, c) - min (a, x, c) l
(0<α<1) a, b, c: adjacent pixels of x (xi,
j-1, xi-1゜JIX11J+1) ) Pcd:C:orrect detect proba
Bility of 1-st errorPmd:M
iss detect probability o
Regarding f no-error method 1, since there is a lot of image quality deterioration due to the probability of false detection, judgment was made for all pixels, and interpolation was performed using a median filter using 3 pixels including the pixel of interest and the pixel 1 below. . On the other hand, method 2 (present invention)
As for, the probability of correct detection is high and the false detection is low.
The same interpolation as in method 1 was performed only for about 15 pixels including pixels determined to be errors.
表3に、エラーコンシールメントにより得られたSNの
結果を表3に示す。この結果、以下のことが明らかにな
った。Table 3 shows the SN results obtained by error concealment. As a result, the following became clear.
(1)画像によって異なるが、方式1.2ともに内挿補
間と組み合わせて画質改善が図られ、S/Nて方式1は
約1〜6dB、方式2(本発明)は約2〜8dBのゲイ
ンがあり、より方式2(本発明)が優れた特性となった
。(1) Although it varies depending on the image, both methods 1 and 2 are combined with interpolation to improve the image quality, and the S/N ratio is approximately 1 to 6 dB for method 1, and approximately 2 to 8 dB for method 2 (the present invention). Therefore, method 2 (the present invention) had better characteristics.
(2)精細度の極めて高い画像を除き、内挿補間により
雑音は気にならない程度に見えにくくなるため、改善さ
れた誤り画累数は表2の誤り検出数とほぼ同じと考えて
よい。このため、方式2(本発明)により視覚的に検知
できる伝送路誤りの数は約15%に減少する。(2) Except for images with extremely high definition, interpolation makes the noise less noticeable to the extent that it is not noticeable, so the improved number of error images can be considered to be approximately the same as the number of detected errors in Table 2. Therefore, the number of transmission path errors that can be visually detected by method 2 (the present invention) is reduced to about 15%.
(3)伝送路誤りの減少量を、本発明を適用した復号化
装置の所要C/Nで換算すると約2dBの運用マージン
の増加になった。(3) When the amount of reduction in transmission path errors is converted to the required C/N of the decoding device to which the present invention is applied, the operating margin increases by about 2 dB.
(4)本発明の方式により画像の平坦部ての改善効果か
特にあり、C/Nの改善以上に視覚的な効果は大きかっ
た。(4) The method of the present invention was particularly effective in improving flat areas of images, and the visual effect was greater than the improvement in C/N.
表3 SNRfor processed pict
ure (dB)Table 3 SNR for processed pic
ure (dB)
第1図(a)は誤り画素とそれを補正する正常画素との
関係を示す側口である。
〕 3
第1図(b)は−次元雑音における水平六回への影響を
示す側口である。
第2図は本発明を適用した復号化装置のブロック図であ
る。
1・・・復号部、 2,3.4・・・遅延部、5
.6.7・・・減算器、8・・・サイン判定部、9、1
0.11・・・絶対値計算部、
12・・・最小値選択部、 13・・・比較判定器、1
4・・・誤り補正部、 15・・・フレームメモリ
、16・・・画素読込部。FIG. 1(a) is a side view showing the relationship between error pixels and normal pixels for correcting them. [3] Figure 1(b) is a side view showing the influence of -dimensional noise on the horizontal six times. FIG. 2 is a block diagram of a decoding device to which the present invention is applied. 1...Decoding unit, 2, 3.4...Delay unit, 5
.. 6.7... Subtractor, 8... Sign determination section, 9, 1
0.11... Absolute value calculation section, 12... Minimum value selection section, 13... Comparison judge, 1
4...Error correction unit, 15...Frame memory, 16...Pixel reading unit.
Claims (1)
複数画素の差分値を計算し、 該差分値の符号比較、および該差分値を量子化特性が有
するエラー伝搬振幅で決定される閾値と比較した結果に
より前記画素Xに伝送誤りが含まれるか否かを見なし検
出し、 誤りが含まれると見なした時は該画素Xの隣接複数画素
から補間値を計算し、 該画素Xの画素値を補間することを特徴とするテレビジ
ョン信号の画質劣化検出方式。[Claims] In a moving image encoding/decoding transmission error correction method, on the decoding side, a difference value between a pixel value of a decoded pixel X and a plurality of adjacent pixels of the pixel X is calculated, and the difference is calculated. Based on the sign comparison of the values and the result of comparing the difference value with a threshold value determined by the error propagation amplitude that the quantization characteristic has, it is determined whether or not the pixel X includes a transmission error, and if an error is included, 1. A method for detecting image quality deterioration of a television signal, comprising: calculating an interpolation value from a plurality of pixels adjacent to the pixel X, and interpolating the pixel value of the pixel X.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP19050590A JP2687693B2 (en) | 1990-07-20 | 1990-07-20 | Television signal quality deterioration detection device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP19050590A JP2687693B2 (en) | 1990-07-20 | 1990-07-20 | Television signal quality deterioration detection device |
Publications (2)
Publication Number | Publication Date |
---|---|
JPH0479586A true JPH0479586A (en) | 1992-03-12 |
JP2687693B2 JP2687693B2 (en) | 1997-12-08 |
Family
ID=16259211
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
JP19050590A Expired - Lifetime JP2687693B2 (en) | 1990-07-20 | 1990-07-20 | Television signal quality deterioration detection device |
Country Status (1)
Country | Link |
---|---|
JP (1) | JP2687693B2 (en) |
-
1990
- 1990-07-20 JP JP19050590A patent/JP2687693B2/en not_active Expired - Lifetime
Also Published As
Publication number | Publication date |
---|---|
JP2687693B2 (en) | 1997-12-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US6327391B1 (en) | Protecting image quality from occurrence of uncorrectable errors | |
US5337088A (en) | Method of correcting an image signal decoded in block units | |
US8902990B2 (en) | Block error compensating apparatus of image frame and method thereof | |
US5347599A (en) | Adaptive interpolation method and apparatus using correlation detection | |
JPH09219863A (en) | Method and device for concealing channel error in video signal | |
CA2090382C (en) | Moving picture receiving method and apparatus | |
EP1396154B1 (en) | Error concealment method and device | |
JP3308675B2 (en) | Encoding device | |
US7039117B2 (en) | Error concealment of video data using texture data recovery | |
US7394855B2 (en) | Error concealing decoding method of intra-frames of compressed videos | |
US20050111743A1 (en) | Error resilient encoding method for inter-frames of compressed videos | |
CN1126376C (en) | Method for eliminating error in bit data flow and apparatus thereof | |
JPH0479586A (en) | Picture quality deterioration detection system for television signal | |
US20030212944A1 (en) | Method and apparatus for error data recovery | |
JP2674241B2 (en) | Video encoding device | |
JP3304415B2 (en) | Decoding device for block transform code | |
JPH0686247A (en) | Receiver/reproducer for digital picture signal | |
JP3759413B2 (en) | Receiving apparatus and receiving method for block transform encoded data | |
JPH0832030B2 (en) | Decoding device for block-encoded image signal | |
JPH0472985A (en) | Frame interpolation system | |
KR100243864B1 (en) | Method and apparatus for concealing channel error of image signals | |
JP3469597B2 (en) | Decoding device for block transform code | |
JP2001078198A (en) | Error concealment control method, coder and image signal transmission system | |
JP3257226B2 (en) | Image processing apparatus and image processing method | |
JP2637976B2 (en) | Block encoding decoding device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
FPAY | Renewal fee payment (event date is renewal date of database) |
Free format text: PAYMENT UNTIL: 20100822 Year of fee payment: 13 |
|
EXPY | Cancellation because of completion of term |