JPH0365770A - Picture processor - Google Patents

Picture processor

Info

Publication number
JPH0365770A
JPH0365770A JP1200205A JP20020589A JPH0365770A JP H0365770 A JPH0365770 A JP H0365770A JP 1200205 A JP1200205 A JP 1200205A JP 20020589 A JP20020589 A JP 20020589A JP H0365770 A JPH0365770 A JP H0365770A
Authority
JP
Japan
Prior art keywords
image
signal
shadow
information
extracted
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.)
Pending
Application number
JP1200205A
Other languages
Japanese (ja)
Inventor
Tetsuo Iijima
飯島 哲生
Kazuaki Yano
和明 矢野
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nippon Telegraph and Telephone Corp
Original Assignee
Nippon Telegraph and Telephone Corp
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Nippon Telegraph and Telephone Corp filed Critical Nippon Telegraph and Telephone Corp
Priority to JP1200205A priority Critical patent/JPH0365770A/en
Publication of JPH0365770A publication Critical patent/JPH0365770A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/20Image preprocessing
    • G06V10/26Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G06V10/273Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion removing elements interfering with the pattern to be recognised

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Character Input (AREA)
  • Image Input (AREA)

Abstract

PURPOSE:To remove the influence of shadow and to obtain information such as characters or patterns with high quality by providing a shadow removing means, which is composed of a peak detection circuit for a picture element density value, peak hold circuit and multiplication arithmetic part, in the front stage of an object information detecting means. CONSTITUTION:In the front stage a means to improve the S/N of normal object information, for example, a Gaussian filter, a shadow removing means 21 composed of the peak detection circuit, peak hold circuit and multiplication arithmetic part is provided in front of the object information detecting means 56 and a background picture with maximum power density is removed at first. Thus, a proper S/N improving means can be obtained to pay attention to information, which are extracted, with the comparatively small power density and the number of times for average processing can be reduced for obtaining the same S/N.

Description

【発明の詳細な説明】 (発明の属する技術分野) 本発明は、陰影部分をもつ画像の中から対象物を確度よ
く抽出する画像処理装置に関する。
DETAILED DESCRIPTION OF THE INVENTION (Technical field to which the invention pertains) The present invention relates to an image processing device that accurately extracts a target object from an image having a shaded area.

(従来の技術) 画像処理の応用分野は広く、(ア)OCR等に代表され
る文字認識、(イ)自律走行ロボッの視覚、(つ)自動
車の登録番号等の文字あるいはパターンの自動認識装置
、などがある。
(Prior art) Image processing has a wide range of applications, including (a) character recognition represented by OCR, (b) vision of autonomous robots, and (d) automatic recognition of characters or patterns such as car registration numbers. ,and so on.

このうち、(ア)のOCR等の装置では照明の品質はあ
る程度保証されているので対象物の認識作業はモデルと
対象物との特徴抽出と判断の問題に帰結できる。この問
題は対象物検出のための環境条件(照明)が整っている
という意味で比較的容易であり、従来から種々の方法が
適用できる。
Among these, since the quality of illumination is guaranteed to some extent in the OCR and other devices (a), the object recognition task can be reduced to a problem of feature extraction and judgment between the model and the object. This problem is relatively easy in the sense that the environmental conditions (illumination) for object detection are in place, and various conventional methods can be applied.

一方、(イ)および(つ)の分野では5文字、パターン
の認識作業は1周囲の環境条件に影響され易く1例えば
明るさ(照明)の均一性は保証されない。
On the other hand, in fields (A) and (T), the work of recognizing 5 characters and patterns is easily influenced by surrounding environmental conditions, and uniformity of brightness (illumination), for example, cannot be guaranteed.

すなわち、被処理画像には、一般に照度のむら、例えば
陰影があるのが普通である。この時、目的とする文字、
パターンの認識率は著しく低下し、ロボットの場合には
自律走行できない。
That is, the image to be processed generally has uneven illuminance, for example, shadows. At this time, the target character,
The pattern recognition rate will drop significantly, and robots will not be able to run autonomously.

このような場合、文字、パターンなど抽出すべき情報(
信号)が陰影情報に較べて一般に高周波数領域にあるこ
とに着目して、微分器、陰影を除去するための高域通過
フィルタ、及び積分器が使われる。エツジ検出のための
高域フィルタ回路はその例である。
In such cases, the information to be extracted such as characters and patterns (
Noting that the signal (signal) is generally in a high frequency range compared to the shadow information, a differentiator, a high-pass filter for removing shadows, and an integrator are used. A high-pass filter circuit for edge detection is an example.

しかし、これらの方法は元々の背景や対象物の照度1反
射率の違いなどのDCレベルの情報を無視しているため
、コントラストが強調されすぎる傾向にあり、実際の対
象物との不一致が生じる。
However, these methods ignore DC level information such as the original background and the difference in illuminance 1 reflectance of the object, so the contrast tends to be overemphasized, resulting in a mismatch with the actual object. .

また、高周波のノイズを強調するため信号のSN比が低
下し、認識率等の低下を招く、または新たな判断手段を
必要とする。
Furthermore, since high-frequency noise is emphasized, the signal-to-noise ratio of the signal decreases, leading to a decrease in recognition rate, etc., or requiring new judgment means.

また、同様な対象物情報を抽出するための手段、例えば
ガウシアンフィルタを使った画像処理が近年行われてい
る。第3図は単純なガウシアンフィルタの機能を説明す
るブロック図であり1図(a)に示すように原画像51
に、比較的信号成分の大きい背景画像と、それに比べる
と小さい被抽出画像を含む場合、原画像51にガウシア
ンノイズ52を重畳させた後、平均化処理回路53を施
した背景画像信号54と、原画像51との差の被抽出画
像信号55(ガウシアンフィルタ出力)をとることによ
り、目的の高周波数信号成分を取り出すことができる。
Furthermore, in recent years, image processing using a means for extracting similar object information, such as a Gaussian filter, has been performed. FIG. 3 is a block diagram explaining the function of a simple Gaussian filter, and as shown in FIG. 1(a), the original image 51
In the case where a background image with relatively large signal components and a relatively small extracted image are included, a background image signal 54 is obtained by superimposing Gaussian noise 52 on the original image 51 and then applying an averaging processing circuit 53, By taking the extracted image signal 55 (Gaussian filter output) that is the difference from the original image 51, the target high frequency signal component can be extracted.

この画像処理回路56では理想的な高域通過フィルタと
見なせる。この過程を走査線に対応する1次元の周波数
領域で表したのが第3図(b)である6図の51〜55
(53は除く)は図(a)に対応している。
This image processing circuit 56 can be regarded as an ideal high-pass filter. Figure 3(b) shows this process in a one-dimensional frequency domain corresponding to the scanning line, 51 to 55 in Figure 6.
(excluding 53) corresponds to figure (a).

この処理において、加えるガウシアンノイズ52のレベ
ルは、被抽出画像信号55より大きく、背景画像信号5
4より小さいことが必要である。このガウシアンノイズ
52は何等かのかたちでそのレベルを判定して与える必
要がある。
In this process, the level of the Gaussian noise 52 to be added is higher than the extracted image signal 55, and the level of the added Gaussian noise 52 is higher than the background image signal 55.
It needs to be smaller than 4. The level of this Gaussian noise 52 needs to be determined and provided in some way.

平均化処理回路53では、1回の処理で背景画像信号5
4とガウシアンノイズ52の比は3dBずつ改善される
。目的の被抽出画像信号55をある値のSN比で処理し
ようとすれば、図(b)のレベル57まで平均化処理を
行う必要がある。即ち、平均化処理回数は理想的な場合
、((52のレベル)−(57のレベル))73回とな
る。
In the averaging processing circuit 53, the background image signal 5 is
4 and Gaussian noise 52 is improved by 3 dB. In order to process the target image signal 55 to be extracted with a certain value of SN ratio, it is necessary to perform the averaging process up to level 57 shown in FIG. 3(b). That is, the number of times of averaging processing is ((52 levels)-(57 levels)) 73 times in an ideal case.

また、この差をとる操作によってSN比も低下する。例
えば、背景画像と被抽出画像の両信号のパワーをそれぞ
れSo、Slとし、5L=0.lX5Oとすると、平均
化後の背景画像信号So′との差が1%ある場合、55
に相当する被抽出画像の誤差はSlの10%となる。即
ち、背景画像と被抽出画像との信号差が大きい場合、被
抽出画像信号のSN比は低下し、後段に再び同様なフィ
ルタを必要とする、等の改善が必要となる。
Moreover, the SN ratio also decreases due to the operation of taking this difference. For example, let the powers of both the background image and extracted image signals be So and Sl, respectively, and 5L=0. Assuming lX5O, if there is a difference of 1% from the averaged background image signal So', then 55
The error of the extracted image corresponding to is 10% of Sl. That is, when the signal difference between the background image and the image to be extracted is large, the SN ratio of the image signal to be extracted decreases, and improvements such as the need for a similar filter at a subsequent stage are required.

(発明が解決しようとする課題) 以上述べたように、従来の2次元画像に対する信号処理
方法ではSN比の改善が必要である。ガウシアンフィル
タに代表される対象物抽出手段を用いる方法でも背景画
像信号が一般に被抽出画像信号より大きいので、大きな
ガウシアンノイズを加えなければならず、それに続く平
均化処理に多くの回数を必要とした。また、ガウシアン
フィルタを多段に構成する必要があった。
(Problems to be Solved by the Invention) As described above, in the conventional signal processing method for two-dimensional images, it is necessary to improve the signal-to-noise ratio. Even in methods that use object extraction means such as Gaussian filters, the background image signal is generally larger than the extracted image signal, so large Gaussian noise must be added, and the subsequent averaging process requires many times. . Furthermore, it was necessary to configure the Gaussian filter in multiple stages.

(発明の目的) 本発明は、陰影部分をもつ2次元画像の中から対象物、
エツジ等を高精度に抽出する画像処理装置を得ることを
目的とする。
(Objective of the Invention) The present invention provides an object to be detected from a two-dimensional image having a shaded area.
The purpose of this invention is to obtain an image processing device that extracts edges etc. with high precision.

(a題を解決するための手段および作用〉本発明は、少
なく画素濃度値のピーク検出回路、ピークホールド回路
、及び乗算演算部から成る陰影除去手段を対象物情報検
出手段の前段に設けたことにより、前記陰影の影響を除
去して高品質な文字、パターンなどの情報を提供できる
(Means and operations for solving problem a) The present invention provides a shadow removal means consisting of at least a pixel density value peak detection circuit, a peak hold circuit, and a multiplication operation section at a stage upstream of the object information detection means. Accordingly, the influence of the shadow can be removed and high quality information such as characters and patterns can be provided.

(実施例) 第1図(a)は1MXNの画素マトリックスからなる画
像の例を示し、1,2.3はその中の数字情報を示す。
(Example) FIG. 1(a) shows an example of an image consisting of a 1MXN pixel matrix, and 1, 2.3 indicate numerical information therein.

背景4−1.4−2は簡単のため同一反射率の白地とす
る。但し、背景4−2の領域は陰影がついている。同図
(b)は、領域5の拡大図を示す。元々の各数字は同じ
反射率(同じ照度の下では同じ輝度、または画像濃度)
であるが、陰影下では数字rlJの上半分6は下半分7
よりも濃度は高い。
Background 4-1.4-2 is a white background with the same reflectance for simplicity. However, the area of the background 4-2 is shaded. FIG. 5B shows an enlarged view of region 5. FIG. Each original number has the same reflectance (same brightness under the same illuminance, or image density)
However, under the shading, the upper half of the number rlJ is 6, and the lower half is 7.
The concentration is higher than that of

同図(c)は、同図(b)の走査線8,9.10の一部
を模式的に示したもので、縦軸は、光強度I (11)
The figure (c) schematically shows a part of the scanning lines 8, 9, and 10 in the figure (b), and the vertical axis is the light intensity I (11).
.

横軸は時間軸tである。 8.8’  ・・・10.1
0’・・・は、走査線に対応する光強度信号の時間変化
である。
The horizontal axis is the time axis t. 8.8'...10.1
0'... is the time change of the light intensity signal corresponding to the scanning line.

領域12から陰影部4−2に入るので、光強度は低くな
る。
Since the light enters the shadow portion 4-2 from the region 12, the light intensity becomes low.

ここで、物体からの光強度工は物体の反射率と照度りの
積であるから9画像入力装置に取り込まれた光強度工は
照度の変化、例えば陰影の影響を顕著に受ける。従って
、照度(陰影)の情報を何等かのかたちで取り込み、物
体の反射率の形で表現すればよい6例えば第1図(C)
で、実際の背景及び文字の反射率は陰影の前後で変わら
ないとしているので領域12で示す陰影の境界の前後で
背景の光強度比から陰影の影響で低下した信号のレベル
を補正できる。
Here, since the light intensity factor from an object is the product of the object's reflectance and illuminance, the light intensity factor taken into the image input device is significantly affected by changes in illuminance, for example, shadows. Therefore, it is only necessary to capture information on illuminance (shading) in some form and express it in the form of reflectance of an object6 For example, Fig. 1 (C)
Since it is assumed that the actual reflectance of the background and characters does not change before and after the shadow, the level of the signal that has decreased due to the influence of the shadow can be corrected from the light intensity ratio of the background before and after the boundary of the shadow shown in area 12.

例えば、その比が1/2になっているとすれば、照度の
変化は陰影の前後の1/2であるので信号13(破線)
が復元できる。これは照度の影響を補正した反射率変化
の信号になっている。この照度変化は、背景の光強度レ
ベル16.16’の変化をとればよいので、ピーク検出
回路とピークホールド回路等で実現できる。ピークホー
ルド時間は注目する被抽出画像信号55の基本波成分に
よって可変できる。ピーク検出及びピークホールド回路
は、ある画素領域における統計情報に基づいて動作する
構成であってもよい1以上の動作は、DCレベルの情報
を保持した理想的な高域通過フィルタと増幅器と考えて
もよい。
For example, if the ratio is 1/2, the change in illuminance is 1/2 before and after the shadow, so signal 13 (dashed line)
can be restored. This is a reflectance change signal that has been corrected for the influence of illuminance. This illuminance change can be realized by a peak detection circuit, a peak hold circuit, etc., since it is sufficient to take a change in the background light intensity level 16.16'. The peak hold time can be varied depending on the fundamental wave component of the extracted image signal 55 of interest. The peak detection and peak hold circuit may be configured to operate based on statistical information in a certain pixel area.One or more operations may be considered as an ideal high-pass filter and amplifier that retains DC level information. Good too.

その後、対象物情報抽出手段、例えばガウシアンフィル
タと平均化処理回路によって対象物情報のSN比を改善
する。これらの処理回路のブロック図を第2図に示す。
Thereafter, the SN ratio of the object information is improved by an object information extraction means, for example, a Gaussian filter and an averaging processing circuit. A block diagram of these processing circuits is shown in FIG.

図において21はピーク検出及びピークホールド回路を
含む陰影処理回路、56は第3図と同様な構成のガウシ
アンフィルタと平均化処理回路でなる画像処理回路であ
る。陰影処理回路21の出力22は背景画像信号の基本
波が除去されているので被抽出画像信号55(SL)が
最大パワーの基本周波数である。従って、ガウシアンフ
ィルタは第2図(b)の23に示すように81より十分
小さいレベルでよく、これに続く画像処理回路56によ
って出力24に示すようにSN比の向上がはかれる。こ
の過程は背景画像の高次の影響をも除去できる。
In the figure, 21 is a shadow processing circuit including a peak detection and peak hold circuit, and 56 is an image processing circuit consisting of a Gaussian filter and an averaging circuit having the same configuration as in FIG. Since the fundamental wave of the background image signal has been removed from the output 22 of the shadow processing circuit 21, the extracted image signal 55 (SL) has the fundamental frequency of the maximum power. Therefore, the level of the Gaussian filter may be sufficiently lower than 81 as shown at 23 in FIG. 2(b), and the subsequent image processing circuit 56 improves the SN ratio as shown at output 24. This process can also remove higher order effects of background images.

次に、抽出すべき画像が2つ(その基本波成分をS2(
<SL)とする)以上の場合、さきに説明したと同様な
処理を施せばよい、複数のパワー(σ□′、σ2′)か
らなるガウシアンフィルタを用意し、S2を抽出する場
合S2>σ2′とし、SlはSL>σ1”>S2とする
ことでSL、S2の画像が得られる。
Next, there are two images to be extracted (their fundamental wave components are S2 (
<SL)) In the above case, the same processing as explained earlier can be performed.If a Gaussian filter consisting of multiple powers (σ□', σ2') is prepared and S2 is extracted, S2>σ2 ', and by setting SL>σ1''>S2, images of SL and S2 can be obtained.

ここで、処理すべき原画像の濃度値は光強度工に対する
以下の変換を施した後先の処理が実行されることに注意
する必要がある。
Here, it must be noted that the density value of the original image to be processed is subjected to the following conversion to the light intensity before the subsequent processing is performed.

画像入力装置に入力する光強度を19画像の濃度をg、
とすると、光電変換の関係より、g=aI’+b   
・・・・・・(1)ここに、γはガンマ定数、bは定数
、また、光強度工は、物体の反射率r、照度をLとする
と、I = r X L   ・・・・・・(2)第1
図(C)で16.16’等は画像の濃度gの値と考えて
よい、陰影情報信号に基づいたIOから13への補正は
(2)式の変換により、照度りを補正して物理的に意味
のある反射率rの差として処理している。
The light intensity input to the image input device is 19. The density of the image is g,
Then, from the relationship of photoelectric conversion, g=aI'+b
...... (1) Here, γ is the gamma constant, b is a constant, and light intensity is the reflectance of the object r, and the illuminance is L, then I = r X L ...・(2) First
In figure (C), 16.16' etc. can be considered as the value of the image density g.The correction from IO to 13 based on the shading information signal is performed by correcting the illuminance and physical This is treated as a meaningful difference in reflectance r.

本発明は以上述べた実施例に限定されない0例えば、前
記ピーク検出回路は何等かの統計情報を基に算出する回
路でよく、またピークホールド回路のピークホールド時
間も処理する画像情報に基づいてもよい。乗算演算部も
ハード/ソフトいずれか、またその両方で実現してもよ
い。
The present invention is not limited to the embodiments described above. For example, the peak detection circuit may be a circuit that calculates based on some statistical information, or the peak hold time of the peak hold circuit may also be based on image information that processes it. good. The multiplication operation unit may also be implemented using hardware, software, or both.

(発明の効果) 以上説明したように、本発明によれば、通常の対象物情
報のSN比を改善する手段、例えばガウシアンフィルタ
の前にピーク検出回路、ピークホールド回路、及び乗算
演算部から成る陰影除去手段を対象物情報検出手段の前
に設け、最大のパワー密度を持つ背景画像をさきに除去
する構成としたので。
(Effects of the Invention) As explained above, according to the present invention, means for improving the SN ratio of normal object information, for example, a peak detection circuit, a peak hold circuit, and a multiplication operation section are provided before a Gaussian filter. The shadow removal means is provided before the object information detection means, and the background image having the maximum power density is removed first.

(a)比較的小さいパワー密度の被抽出情報に注目した
適正なSN比改善手段がとれる。
(a) Appropriate means for improving the SN ratio can be taken by focusing on information to be extracted with relatively low power density.

(b)同−SN比を得るのに平均化処理の回数が少なく
できる。
(b) The number of averaging processes can be reduced to obtain the same SN ratio.

等のメリットがある。There are other benefits.

【図面の簡単な説明】[Brief explanation of drawings]

第1図および第2図は本発明の一実施例を示す図、第3
図は単純なガウンアンフィルタの機能を説明する図であ
る。 1.2,3  ・・・MXN画素マトリックスの中の数
字情報、4−1.4−2 ・・・背景、 8,9.10
・・・模式的に示す走査線の一部、 8,8′  ・・
・10.10’ ・・・走査線に対応する光強度信号の
時間変化、12・・・陰影の境界領域、13・・・信号
10からの光強度レベル16.16′の変化に対して復
元された信号、21・・・ ピーク検出回路とピークホ
ールド回路を含む陰影処理回路、24・・・出力信号、
51・・・原画像、52・・・ガウシアンノイズ、53
・・・平均化処理回路、54・・・背景画像信号、55
・・・被抽出画像信号(ガウシアンフィルタ出力)、5
6・・・ガウシアンフィルタと平均化処理回路からなる
画像処理回路。
1 and 2 are diagrams showing one embodiment of the present invention, and FIG.
The figure is a diagram illustrating the function of a simple Gaun filter. 1.2, 3...Numeric information in MXN pixel matrix, 4-1.4-2...Background, 8,9.10
...Part of the scanning line schematically shown, 8,8'...
・10.10'...Time change of light intensity signal corresponding to the scanning line, 12...Shadow boundary area, 13...Restored for change in light intensity level 16.16' from signal 10 signal, 21... shading processing circuit including a peak detection circuit and a peak hold circuit, 24... output signal,
51... Original image, 52... Gaussian noise, 53
... Averaging processing circuit, 54 ... Background image signal, 55
...Extracted image signal (Gaussian filter output), 5
6... Image processing circuit consisting of a Gaussian filter and an averaging processing circuit.

Claims (1)

【特許請求の範囲】[Claims] 画像の中から対象物を抽出する画像処理装置において、
少なくとも画素濃度値のピーク検出回路、ピークホール
ド回路、及び乗算演算部から成る陰影除去手段を対象物
情報検出手段の前段に設けたことを特徴とする画像処理
装置。
In an image processing device that extracts a target object from an image,
An image processing apparatus characterized in that a shadow removing means comprising at least a pixel density value peak detection circuit, a peak hold circuit, and a multiplication operation section is provided at a stage upstream of the object information detection means.
JP1200205A 1989-08-03 1989-08-03 Picture processor Pending JPH0365770A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP1200205A JPH0365770A (en) 1989-08-03 1989-08-03 Picture processor

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP1200205A JPH0365770A (en) 1989-08-03 1989-08-03 Picture processor

Publications (1)

Publication Number Publication Date
JPH0365770A true JPH0365770A (en) 1991-03-20

Family

ID=16420552

Family Applications (1)

Application Number Title Priority Date Filing Date
JP1200205A Pending JPH0365770A (en) 1989-08-03 1989-08-03 Picture processor

Country Status (1)

Country Link
JP (1) JPH0365770A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2016024694A (en) * 2014-07-23 2016-02-08 富士通株式会社 Image processing apparatus, image processing method, and program

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2016024694A (en) * 2014-07-23 2016-02-08 富士通株式会社 Image processing apparatus, image processing method, and program

Similar Documents

Publication Publication Date Title
US6269194B1 (en) Method and apparatus using image subtraction and dynamic thresholding
CN101980521B (en) Image sharpening method and related device
US8306352B2 (en) Image processing method and apparatus
KR20220028066A (en) A method, computer program product and computer readable medium for generating a mask for a camera stream
JP2004038885A (en) Image feature learning type defect detection method, defect detection device and defect detection program
US20090257673A1 (en) Method and system for noise level detection in image data
US8090215B2 (en) Method and system for noise level detection in image data
CN111091107A (en) Face region edge detection method and device and storage medium
US8208753B2 (en) Method and system for noise level detection in image data
CN111311610A (en) Image segmentation method and terminal equipment
JPH0365770A (en) Picture processor
JPH0981754A (en) Movement detecting circuit
CN113256534B (en) Image enhancement method, device and medium
CN113222978A (en) Target identification method
JPS62138986A (en) Character recognizing device
JPH04220885A (en) Background elimination method and its execution device
JP2001228050A (en) Apparatus and method for detection of point defect
JPS63115279A (en) Area extracting system
JPH02206882A (en) Picture processor
JP2003346165A (en) Detection method of work boundary
JPH08202870A (en) Picture processing method
JPH02128575A (en) Picture processor
JPH05165963A (en) Extraction of image contour
JPH04278785A (en) Background eliminating method and device for excution thereof
JP3055168B2 (en) Image inspection equipment