JP3706334B2 - Image capturing device, image capturing method, fingerprint verification device, and fingerprint verification method - Google Patents

Description

【００７９】
縦方向の一列の長さはｍピクセルであるので、Ｐピクセル当たりの隆線数に規格化するためにＰ／ｍを数１に乗ずることで、以下の数２に示すように、ピクセル数Ｐ当たりの横方向の平均隆線数Ｎｖを算出する。
【００８０】
【数２】

【００８１】
縦方向についても同様にして求めると、ピクセル数Ｐ当たりの縦方向の平均隆線数Ｎｈは、以下に示す数３で表される。
【００８２】
【数３】

【００８３】
したがって、隆線数指標Ｎは、Ｎ＝ＭＡＸ（Ｎｖ，Ｎｈ）と表され、ＮｖとＮｈのうち大きい方を隆線数指標Ｎとする。
【００８４】
図１６は、隆線判定条件を説明する図である。図１６において、縦軸は各ピクセルの階調値、横軸は判定エリアの縦または横方向のピクセルを表している。各ピクセルの階調値は、指紋画像の凹凸により波打っており、この波の数が隆線の数となる。隆線の数え方は、例えば、２つの閾値ＨＶＴＨとＬＶＴＨを用いて、ＨＶＴＨからＬＶＴＨをまたいだ場合に隆線が１つあるとカウントする。ここで、２つの閾値を用いるのは、ノイズ等による画素を隆線としてカウントするのを防ぐためであるが、もちろんＨＶＴＨ＝ＬＶＴＨとして閾値を１つにすることもできる。
【００８５】
また、指紋画像の階調値を平滑化してから隆線を数えることもできる。このようにすると、ノイズや装置のばらつき等による指紋画像の変動による影響を低減させることができる。
【００８６】
隆線数指標Ｎは、この値が極端に小さい場合は、指紋画像がつぶれてしまっている状態である。逆に、隆線数指標Ｎの値が大きい場合は、指紋画像がかすれて、隆線がとぎれてしまっている状態である。したがって、隆線数指標Ｎが極端に小さい場合には、画像がつぶれないように画質を薄めにするように、逆に隆線数指標Ｎが極端に大きい場合には、画像を濃くするように、パラメータＡ、Ｂを調整する。このようにすることにより、指紋パターンの濃淡に左右されない適正な指紋画像を取得することができる。
【００８７】
なお、隆線（図１５の指紋画像において黒い線）の代わりに谷線（図１５の指紋画像において白い線）から隆線数指標を算出することができる。この場合も、隆線数指標と同等の評価指標となる。谷線の数え方は、例えば図１６で説明すると、２つの閾値ＨＶＴＨとＬＶＴＨにおいて、ＬＶＴＨからＨＶＴＨをまたいだ場合に谷線が１つあると数えることができる。また、この場合も閾値を１つにすることができる。
【００８８】
本実施の形態においても、装置の構成は、前述した他の実施の形態と同様の画像取り込み装置（指紋照合装置）と同様であり、図１７に示すように取り込み制御部３ａが動作するものである。以下、本実施の形態の指紋照合装置における取り込み制御部３ａの動作について説明する。なお、図９と同等の構成用件については、説明を適宜省略する。
【００８９】
まず、指置き指示部６から照合を行う利用者に対して画像取り込み部１の検出面１２（図２）への指置きを指示する（ステップＳ１）。指置き指示の具体例としては、ＬＥＤなどの発光素子による表示器を用いて指示を表示する形態や、所定の指示ＯＮや音声メッセージなどを出力する形態がある。指置き指示により利用者が指を検出面１２へ置いたことを指置き検出部７で検出した後（ステップＳ２）、画像取り込み部１が指紋画像の取り込み、画像データ２を出力する（ステップＳ３）。
【００９０】
画像データ２が出力されると、取り込み制御部３ａは、画像データ２の画質評価として、まず、取り込んだ画像の隆線数指標Ｎを算出する（ステップＳ２０１）。なお、隆線数指標Ｎは、取り込んだ画像の全画素のみならず、取り込んだ画像の所定領域、例えば検出面１２の中央付近に設けられた領域から算出するようにしてもよい。この場合は、全画素を対象とする場合より短い時間で隆線数指標Ｎを算出することができる。
【００９１】
つぎのステップＳ２０２〜ステップＳ２１２で、取り込み制御部３ａは、上述のことにより得られた隆線数指標Ｎと、予め設定してある閾値Ｎ_th1とＮ_th2とを比較する。
まず、取り込み制御部３ａは、ステップＳ２０２で、算出した隆線数指標ＮがＮ_th2以下と判断した場合、ステップＳ２０３に進み、今度は、隆線数指標Ｎが閾値Ｎ_th1以上かどうかを判断する。ステップＳ２０３で、隆線数指標Ｎが、閾値Ｎ_th1以上と判断した場合、ステップＳ２０４に進み、取り込み制御部３ａは、取り込んだ画像データを照合手段３ｂに出力する。
【００９２】
一方、ステップＳ２０２で、隆線数指標Ｎが閾値Ｎ_th2を超えていると判断した場合、ステップＳ２０５に進み、まず、パラメータ設定部１ａに設定されているパラメータ値Ｂが、最大値であるかどうかを判断する。
ステップＳ２０５で、設定されているパラメータ値Ｂが最大値でないと判断した場合、ステップＳ２０６に進み、取り込み制御部３ａは、パラメータ設定部１ａにおけるパラメータ値Ｂを所定の値だけ増加させ、ステップＳ３に戻る。また、ステップＳ２０５で、パラメータ設定部１ａに設定されているパラメータ値Ｂが最大値であると判断した場合、ステップＳ２０７に進み、パラメータ設定部１ａに設定されているパラメータ値Ａが最大値であるかどうかを判断する。
【００９３】
ステップＳ２０７で、設定されているパラメータ値Ａが最大値でないと判断した場合、ステップＳ２０８に進み、取り込み制御部３ａは、パラメータ設定部１ａにおけるパラメータ値Ａを所定の値だけ増加させ、ステップＳ３に戻る。ステップＳ２０７で、最大値であると判断したら、ステップＳ２０４に進み、取り込み制御部３ａは、取り込んだ画像データを照合手段３ｂに出力する。
Ｎ＞Ｎ_th2の場合は、取り込んだ画像がかすれて薄すぎる状態であるため、以上のことにより、パラメータ値Ａ，Ｂを増加させて再度指紋画像の取り込みを行う。
【００９４】
ステップＳ２０３の判断で、隆線数指標Ｎが閾値Ｎ_th1未満と判断した場合、ステップＳ２０９に進み、ここでは、パラメータ設定部１ａに設定されているパラメータ値Ａが、最小値例えば０であるかどうかを判断する。
設定されているパラメータ値Ａが０でないと判断した場合、ステップＳ２１０に進み、取り込み制御部３ａは、パラメータ設定部１ａにおけるパラメータ値Ａを所定の値だけ減少させ、ステップＳ３に戻る。また、ステップＳ２０９で、パラメータ設定部１ａに設定されているパラメータ値Ａが０であると判断した場合、ステップＳ２１１に進み、パラメータ設定部１ａに設定されているパラメータ値Ｂが最小値であるかどうかを判断する。
【００９５】
ステップＳ２１１で、設定されているパラメータ値Ｂが最小値でないと判断した場合、ステップＳ２１２に進み、取り込み制御部３ａは、パラメータ設定部１ａにおけるパラメータ値Ｂを所定の値だけ減少させ、ステップＳ３に戻る。ステップＳ２１１で、最小値であると判断したら、ステップＳ２０４に進み、取り込み制御部３ａは、取り込んだ画像データを照合手段３ｂに出力する。
Ｎ＜Ｎ_th1の場合は、取り込んだ画像がつぶれている状態であるため、以上のことにより、パラメータ値Ａ，Ｂを減小させて再度指紋画像の取り込みを行うようにしている。
【００９６】
取り込み制御部３ａは、画像データにおける画質が、所定の条件を満たすまで、上記一連の動作を繰り返し、条件を満たした場合に、認証に適した画像として画像データを照合手段３ｂに出力する。このように、本実施の形態によれば、指紋パターンの濃淡および個人差や状態変化に影響されずに、つぶれまたは擦れがない指紋画像を取得することができる。
なお、例外処理として、パラメータＡ、Ｂの値が最大値または最小値になっても条件を満たさない場合は、この時の取り込み画像を出力する。
本実施の形態において、指紋の判定エリアは、必要に応じて場所および大きさを自由に設定することができる。
【００９７】
＜実施の形態４＞
本実施の形態は、画像処理装置および指紋照合装置における評価指標に、ヒストグラム指標と隆線数指標とを、組み合わせて適用したものである。
例えば、ヒストグラム指標と隆線数指標の各々について所定の範囲内となるようにパラメータＡ、Ｂを変更するようにしてもよい。
また、ヒストグラム指標と隆線数指標から所定の演算により新たな指標を導出し、この指標に基づいてパラメータＡ、Ｂを変更するようにしてもよい。
【００９８】
このような構成にすることより、指紋パターンの濃淡および個人差や状態変化に影響されずに、コントラストが強くかつつぶれまたは擦れがない指紋画像を取得することができる。
なお、本実施の形態において、隆線数指標を第２の実施の形態で説明した谷線を用いて算出した隆線数指標に置き換えて適用することも可能である。
【００９９】
また、上述した実施の形態では、アナログ値を２５６階調のデジタル値に変換する場合について説明したが、２５６以外の階調に変換することも可能である。
本発明は、パラメータ設定部１ａに設定され、画像取り込み部１における変換に用いられるパラメータとして、パラメータ値Ａとパラメータ値Ｂを用いるようにしているが、これに限るものではない。例えば、輝度と分解能を制御するために、信号の変換領域の最小値と最大値とを、各々パラメータ値Ａ，パラメータ値Ｂとして用いるようにしても同様である。また、設定されるパラメータの数もこれに限定されず、必要に応じて自由に変更することができる。
【０１００】
なお、上記実施の形態では、評価指標として、画像データの濃淡度合いを示す濃淡指標と画像の空間周波数成分を示す指標とを組み合わせたもの、および、画像の濃淡を表現したヒストグラムから生成されるヒストグラム指標と画像中の隆線の数に基づいて生成される隆線数指標とを組み合わせたものを例示したが、これらの組み合わせに限るものではない。各評価指標を適宜に組み合わせて新たな評価指標として用いるようにしても良いことは、いうまでもない。
【０１０１】
【発明の効果】
以上説明したように、本発明によれば、例えば指紋照合の対象として、ある範囲内に画像の評価指標が入る画像データを取り込むようにしたので、所望とする状態の画像が得られるようになるという優れた効果が得られ、例えば、指紋照合の精度を向上させることができるようになる。
【図面の簡単な説明】
【図１】 本発明の実施の形態における画像取り込み装置を用いた指紋照合装置の構成を示す構成図である。
【図２】 図１の指紋照合装置の画像取り込み部の一部構成を示す斜視図である。
【図３】 画像取り込み部に取り込まれた指紋の画像データの状態を示す説明図である。
【図４】 画像取り込み部の概略的な構成を示す構成図である。
【図５】 Ａ／Ｄ変換回路１４に設定されるパラメータの一例を示す説明図である。
【図６】 本発明の実施の形態における指紋照合装置の動作を示すフローチャートである。
【図７】 取り込んだ画像データと、この画像データが示す画像の濃淡を表現したヒストグラムを示す説明図である。
【図８】 取り込んだ画像データと、この画像データが示す画像の濃淡を表現したヒストグラムを示す説明図である。
【図９】 本発明の他の形態における指紋照合装置の動作を示すフローチャートである。
【図１０】 本発明の他の形態における指紋照合装置の一部構成を示す構成図である。
【図１１】 本発明の他の形態における指紋照合装置の一部構成を示す構成図である。
【図１２】 本発明の他の形態における指紋照合装置の一部構成を示す構成図である。
【図１３】 本発明の他の形態における指紋照合装置の一部構成を示す回路図である。
【図１４】 図１３に示す指紋照合装置の具体的な動作例を示すタイミングチャートである。
【図１５】 隆線指標を説明する説明図である。
【図１６】 隆線判定条件を説明する特性図である。
【図１７】 本発明の他の形態における指紋照合装置の動作を示すフローチャートである。
【符号の説明】
１…画像取り込み部、１ａ…パラメータ設定部、２…画像データ、３…制御部、３ａ…取り込み制御部、３ｂ…照合手段、４…記憶部、５…照合結果、６…指置き指示部、７…指置き検出部、Ｇ₁〜Ｇ_N…登録画像データ。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image capturing device and an image capturing method that capture the shape of a target object as image data, and a fingerprint collation device and fingerprint collation method that perform collation by detecting irregularities due to fingerprints as image data using these. .
[0002]
[Prior art]
When performing fingerprint collation, in order to improve collation accuracy, it is necessary to capture a fingerprint image so that the unevenness of the fingerprint is clear. As a fingerprint sensor used to capture a fingerprint image with this type of fingerprint collation device, there are an optical fingerprint sensor, a capacitive fingerprint sensor, and the like. The fingerprint sensor detects unevenness of a fingerprint as a difference in optical refractive index or a difference in capacitance value, and generates a grayscale image.
[0003]
When the fingerprint verification apparatus authenticates a user using a fingerprint, first, the fingerprint data of the user is registered in advance. In this state, an image of the user's fingerprint is acquired by the fingerprint sensor, fingerprint data is generated, the generated fingerprint data is compared with the registered fingerprint data, and if the two match, the user is identified. recognize.
[0004]
[Problems to be solved by the invention]
By the way, human fingertips have individual differences in skin conditions such as dry skin and oily skin. Moreover, even if it is the same person, the skin state of a fingertip changes with seasons and physical condition. The conventional fingerprint collation device has a problem that an accurate fingerprint image necessary for collation cannot be obtained because a fingerprint image that changes as described above is detected under the same conditions. Thus, if an image in a desired state cannot be obtained, accurate fingerprint matching is hindered.
[0005]
The present invention has been made to solve the above-described problems, and an object thereof is to obtain an image in a desired state.
[0006]
[Means for Solving the Problems]
An image capturing device according to an aspect of the present invention converts an object shape into an electrical quantity according to a parameter value set in a parameter setting unit, and outputs image data representing an image corresponding to the object shape The image capturing unit and the image data output from the image capturing unit are received, an evaluation index for evaluating the image quality of the image is calculated from the image data, and if the evaluation index is out of the preset reference value range For example, the parameter value set in the parameter setting unit is changed so as to be within the range of the reference value, and the capture control unit that outputs the image data received from the image capture unit whose evaluation index is within the range of the reference value; It is equipped with.
According to this image capturing device, the conversion state in the image capturing unit is changed according to the state of the image data output from the image capturing unit.
[0007]
In the image capturing device, when the parameter value set in the parameter setting unit is changed, the image capturing unit outputs image data converted again according to the changed parameter value. Further, in the above invention, the image capturing unit includes a detection element that converts the shape of the object into an analog signal, and an analog signal output from the detection element according to the parameter value set in the parameter setting unit. A parameter value set in the parameter setting unit is a conversion range and conversion resolution when converting an analog signal into a digital signal.
[0009]
In the image capturing device, the evaluation index calculated by the capturing control unit may be, for example, a histogram index generated from a histogram representing the density of an image. In this case, the capture control unit obtains the maximum value on the gradation side where the density of the image is dark in the histogram, and the minimum value closest to the maximum value on the gradation side where the density of the image is lighter than the gradation value indicating the maximum value. The ratio may be calculated as a histogram index.
[0010]
In the image capturing device, the evaluation index calculated by the capturing control unit may be, for example, a ridge number index generated based on the number of ridges in the image. In this case, the capture control unit performs the horizontal average ridge number per unit length in the horizontal direction of the image and the vertical direction that is the average ridge number per unit length in the vertical direction of the image. The number of ridges may be obtained, and the larger of the average number of ridges in the horizontal direction and the number of ridges in the vertical direction may be calculated as the ridge number index.
The evaluation index may be a combination of a histogram index generated from a histogram representing the density of an image and a ridge number index generated based on the number of ridges in the image.
[0011]
An image capturing method according to an aspect of the present invention generates image data representing an image corresponding to a shape of an object by converting the shape of the object into an electrical signal according to a preset parameter value. An evaluation index for evaluating the image quality of the image is calculated, and the parameter value is changed so that the evaluation index falls within a preset reference value range.
According to this image capturing method, the conversion state into image data is changed depending on the state of the generated image data.
[0012]
  In the image capturing method, the parameter value is a conversion range and conversion resolution when converting an analog signal to a digital signal,fingerprintIs converted into an analog signal, and the analog signal is converted into a digital signal according to the parameter value and output as image data..
[0013]
In the image capturing method, the evaluation index may be, for example, a histogram index generated from a histogram expressing the density of an image, or a ridge number index generated based on the number of ridges in the image. Good. For example, the histogram index includes a maximum value on the gradation side where the density of the image is dark in the histogram and a minimum value closest to the maximum value on the gradation side where the density of the image is lighter than the gradation value indicating the maximum value. A ratio may be used. Further, for example, the average number of ridges that are the average number of ridges per unit length in the horizontal direction of the image, and the number of vertical ridges that are the average number of ridges per unit length in the vertical direction of the image, And the larger of the number of horizontal average ridges and the number of vertical ridges may be used as the ridge number index.
The evaluation index may be a combination of a histogram index generated from a histogram representing the density of an image and a ridge number index generated based on the number of ridges in the image.
[0014]
According to an aspect of the present invention, there is provided a fingerprint collation device that converts an unevenness due to a fingerprint into an electrical quantity according to a parameter value set in a parameter setting unit, and outputs image data representing a fingerprint image corresponding to the unevenness of the fingerprint Image data output from the image capturing unit and the image capturing unit, and an evaluation index for evaluating the image quality of the fingerprint image is calculated from the image data. If the evaluation index is outside the preset reference value range For example, the parameter value set in the parameter setting unit is changed so as to be within the range of the reference value, and the capture control unit that outputs the image data received from the image capture unit whose evaluation index is within the range of the reference value; , Provided with collation means for comparing and collating image data output from the capture control unit with registered image data prepared in advance A.
According to this fingerprint collation device, the conversion state in the image capturing unit is changed according to the state of the image data output from the image capturing unit.
[0015]
In the fingerprint collation apparatus, when the parameter value set in the parameter setting unit is changed, the image capturing unit outputs image data converted again according to the changed parameter value.
In the fingerprint collation device, the image capturing unit digitally converts the detection element for converting the irregularities caused by the fingerprint into an analog signal and the analog signal output from the detection element according to the parameter value set in the parameter setting unit. A parameter value set in the parameter setting unit is composed of an A / D conversion circuit that converts the signal into a signal and outputs it as image data. The parameter value is a conversion range and conversion resolution for converting an analog signal into a digital signal.
[0017]
In the fingerprint collation apparatus, the evaluation index calculated by the capture control unit may be, for example, a histogram index generated from a histogram representing the density of an image. In this case, the capture control unit obtains the maximum value on the gradation side where the density of the image is dark in the histogram, and the minimum value closest to the maximum value on the gradation side where the density of the image is lighter than the gradation value indicating the maximum value. The ratio may be calculated as a histogram index.
[0018]
In the fingerprint collation device, the evaluation index calculated by the capture control unit may be, for example, a ridge number index generated based on the number of ridges in the image. In this case, the capture control unit performs the horizontal average ridge number per unit length in the horizontal direction of the image and the vertical direction that is the average ridge number per unit length in the vertical direction of the image. The number of ridges may be obtained, and the larger of the average number of ridges in the horizontal direction and the number of ridges in the vertical direction may be calculated as the ridge number index.
The evaluation index may be a combination of a histogram index generated from a histogram representing the density of an image and a ridge number index generated based on the number of ridges in the image.
[0019]
The fingerprint collation device further includes a finger placement detection unit that detects that a finger is placed on the image capture unit, and the image capture unit detects that the finger placement detection unit has placed a finger on the image capture unit. Then, the unevenness due to the fingerprint may be converted into an electric quantity according to the parameter value set in the parameter setting unit, and image data representing the fingerprint image corresponding to the unevenness of the fingerprint may be output.
[0020]
The fingerprint collation method according to one aspect of the present invention generates image data representing a fingerprint image corresponding to the fingerprint unevenness by converting the unevenness due to the fingerprint into an electric signal according to a preset parameter value. An image for which an evaluation index for evaluating the image quality is calculated from the data, the parameter value is changed so that the evaluation index is within a preset reference value range, and the evaluation index is within the reference value range. The comparison is made by comparing the data with registered image data prepared in advance.
According to this fingerprint collation method, the conversion state into image data is changed depending on the state of the generated image data.
[0021]
  In the fingerprint collation method, the parameter value is a conversion range and conversion resolution when an analog signal is converted into a digital signal, and irregularities due to fingerprints are converted into an analog signal, and the analog signal is converted into a digital signal according to the parameter value. It is an attempt to convert and output as image data.
[0022]
In the fingerprint collation method, the evaluation index may be, for example, a histogram index generated from a histogram expressing the density of an image or a ridge number index generated based on the number of ridges in the image. Good.
For example, the histogram index includes a maximum value on the gradation side where the density of the image is dark in the histogram and a minimum value closest to the maximum value on the gradation side where the density of the image is lighter than the gradation value indicating the maximum value. A ratio may be used. Further, for example, the average number of ridges that are the average number of ridges per unit length in the horizontal direction of the image, and the number of vertical ridges that are the average number of ridges per unit length in the vertical direction of the image, And the larger of the number of horizontal average ridges and the number of vertical ridges may be used as the ridge number index.
In the fingerprint collation method, the evaluation index is a combination of a histogram index generated from a histogram representing the density of an image and a ridge number index generated based on the number of ridges in the image. May be.
[0023]
DETAILED DESCRIPTION OF THE INVENTION
  Hereinafter, embodiments of the present invention will be described with reference to the drawings.. Figure1 is a configuration diagram illustrating a configuration of an image capturing device according to an embodiment of the present invention. In this embodiment, the case where the image capturing device is applied to fingerprint collation is taken as an example, and the following description will be made as a fingerprint collating device. The fingerprint collation apparatus includes an image capturing unit 1 that first converts an uneven image of a fingerprint into image data that is an electrical signal and outputs the image data. The image capturing unit 1 determines the conversion state into image data based on the parameters set in the parameter setting unit 1a.
[0024]
Further, the apparatus of FIG. 1 has registered image data G for fingerprint verification.₁~ G_NStorage unit 4 for storing, finger placement instruction unit 6 for instructing a user performing collation to place a finger on the image capturing unit 1, and finger placement detection for detecting that the finger is placed on the image capturing unit 1 And a control unit 3 for controlling the entire apparatus.
[0025]
The control unit 3 includes a CPU that performs predetermined arithmetic processing by a program. First, the control unit 3 includes capture control means 3a that determines the image data output from the image capture unit 1 and changes the parameters of the parameter setting unit 1a. . The control unit 3 also includes image data 2 captured by the image capturing unit 1 and registered image data G stored in the storage unit 4.₁~ G_NAnd collation means 3b for outputting the result as a collation result 5. These capture control means 3a and collation means 3b are realized by a program.
[0026]
FIG. 2 is a perspective view illustrating a configuration of the image capturing unit 1. On the detection surface 12 of the image capturing unit 1, a large number of sensor cells 11 are arranged in a matrix in the vertical and horizontal directions. The sensor cell 11 is composed of an element that converts a minute unevenness of an object into an electric quantity, and the unevenness of the fingerprint 22 is detected by each sensor cell 11 when the finger 21 contacts the detection surface 12. As a detection result of all sensor cells, one image data 2 representing a fingerprint image as shown in FIG.
[0027]
FIG. 4 is a configuration diagram showing the configuration of the image capturing unit 1. The sensor cell 11 includes a detection element 11a and a sensor circuit 11b. The detection element 11a is, for example, a capacitive sensor composed of an electrode having an insulating layer on the surface. In addition, the detection element 11a may be an optical sensor configured from, for example, a photodiode.
[0028]
The detection element 11a converts the unevenness of the fingerprint into an electrical signal, and the converted signal is amplified by the sensor circuit 11b. The output of the sensor circuit 11 b of each sensor cell 11 is connected to a common data line 13. The sensor cells 11 are sequentially selected and output an analog signal corresponding to the unevenness of the fingerprint to the data line 13.
The A / D conversion circuit 14 sequentially converts an analog signal transmitted via the data line 13 into, for example, a digital signal of 256 gradations and outputs it according to the parameters set in the parameter setting unit 1a. When the digital signals of all the sensor cells 11 are arranged in a matrix reflecting the arrangement of each sensor cell, fingerprint image data in a state as shown in FIG. 3 is obtained.
[0029]
The A / D conversion circuit 14 converts an analog signal corresponding to the unevenness of the fingerprint into a digital signal. Correspondence between the analog value and the digital value is determined according to the parameter value set in the parameter setting unit 1a. The relationship changes.
FIG. 5 is an explanatory diagram showing an example of parameters set in the A / D conversion circuit 14 (FIG. 4). Parameters A and B are each a rational number of 0 or more. If the analog value input to the A / D conversion circuit 14 is a level lower than the set parameter value (A + B), fingerprint image data of gradation 0 is obtained from the A / D conversion circuit 14. The fingerprint image data of gradation 0 is a black image.
[0030]
If the analog value input to the A / D conversion circuit 14 is greater than or equal to the set parameter value (A + 255 × B), fingerprint image data of gradation 255 is obtained from the A / D conversion circuit 14. The fingerprint image data of gradation 255 is a white image. Further, if the analog value input to the A / D conversion circuit 14 is at a level that is not less than the parameter value (A + n × B) and less than the parameter value {A + (n + 1) × B}, the A / D conversion circuit 14 To obtain image data of gradation n.
[0031]
As described above, when the parameter value A is set large, the fingerprint image obtained from the A / D conversion circuit 14 becomes black, and when the parameter value A is set small, the fingerprint image obtained from the A / D conversion circuit 14 becomes white. Become. If the parameter value B is set large, the resolution of the fingerprint image obtained from the A / D conversion circuit becomes coarse, and if the parameter value B is set small, the resolution of the fingerprint image obtained from the A / D conversion circuit 14 becomes fine. . The number of gradations is not limited to 256, and may be 64 or 128.
[0032]
Although the parameter value A and the parameter value B described above are used as parameters set in the parameter setting unit 1a and used for conversion in the image capturing unit 1, the present invention is not limited to this. For example, the same applies to the case where the minimum value and maximum value of the signal conversion region are used as the parameter value A and the parameter value B, respectively, in order to control the luminance and resolution.
[0033]
As described above, the brightness and resolution of the fingerprint image obtained from the A / D conversion circuit 14 can be changed by changing the parameter value of the A / D conversion circuit 14. Therefore, when detecting fingerprint images of human fingertips, such as dry fingers and oily skin, whose state changes depending on individual differences, seasons and physical conditions, it is necessary to set the appropriate parameter values and capture the images. Therefore, it is possible to acquire an accurate fingerprint image that is not affected by individual differences or state changes.
[0034]
  Next, using the flowchart of FIG.,fingerThe operation of the pattern matching device (image capturing device) will be described.
  First, the capture control unit 3a of the control unit 3 instructs the user who performs collation from the finger placement instruction unit 6 to place the finger on the detection surface 12 (FIG. 2) of the image capture unit 1 (step S1). As specific examples of the finger placement instruction, there are a form in which an instruction is displayed using a display device using a light emitting element such as an LED, and a form in which a predetermined instruction ON, a voice message, or the like is output. After the finger placement detection unit 7 detects that the user has placed the finger on the detection surface 12 according to the finger placement instruction (step S2), the image capture unit 1 captures the fingerprint image and outputs the image data 2 (step S3). ).
[0035]
When the image data 2 is output, the capture control unit 3a first calculates a shade index Y indicating the shade level of the captured image as an image quality evaluation of the image data 2 (step S4). Any index may be used as long as it indicates the density balance of the image data to be evaluated. For example, the ratio of the number of pixels that are equal to or lower than a predetermined gradation value among all the pixels of the image data may be used as an index. In addition, a predetermined area, for example, an area provided near the center of the detection surface 12 may be used as an index calculation target area without using the total number of pixels as an index calculation target. In comparison, the shade index can be calculated in a short time.
[0036]
In the next step S5 to step S15, the capture control unit 3a determines the tone index Y obtained as described above and the preset threshold value Y._th1And Y_th2And compare.
First, the capture control unit 3a determines that the calculated shade index Y is the threshold value Y in step S5._th1Judge whether it is above. In step S5, the shade index Y is the threshold value Y._th1If it is determined as above, the process proceeds to step S6, and this time the density index Y is set to the threshold value Y._th2Determine if: In step S6, the shade index Y is changed to the threshold value Y._th2If it is determined as follows, the process proceeds to step S7, and the capture control unit 3a outputs the captured image data to the matching unit 3b.
[0037]
On the other hand, in step S5, the shade index Y is a threshold value Y._th1When it is determined that the value is less than the value, the process proceeds to step S8, and first, it is determined whether or not the parameter value B set in the parameter setting unit 1a is the maximum value.
If it is determined in step S8 that the set parameter value B is not the maximum value, the process proceeds to step S9, and the capture control unit 3a increases the parameter value B in the parameter setting unit 1a by a predetermined value, and then proceeds to step S3. Return. If it is determined in step S8 that the parameter value B set in the parameter setting unit 1a is the maximum value, the process proceeds to step S10, and the parameter value A set in the parameter setting unit 1a is the maximum value. Determine whether or not.
[0038]
If it is determined in step S10 that the set parameter value A is not the maximum value, the process proceeds to step S11, and the capture control unit 3a increases the parameter value A in the parameter setting unit 1a by a predetermined value, and then proceeds to step S3. Return. If it is determined in step S10 that the value is the maximum value, the process proceeds to step S7, and the capture control unit 3a outputs the captured image data to the matching unit 3b.
Y <Y_th1In this case, since the captured image is in a slightly thin state, the fingerprint values are captured again by increasing the parameter values A and B as described above.
[0039]
As a result of the determination in step S6, the shade index Y is the threshold value Y._th2If it is determined that the value exceeds the value, the process proceeds to step S12, where it is determined whether or not the parameter value A set in the parameter setting unit 1a is a minimum value, for example, 0.
If it is determined that the set parameter value A is not 0, the process proceeds to step S13, and the capture control unit 3a decreases the parameter value A in the parameter setting unit 1a by a predetermined value, and returns to step S3. If it is determined in step S12 that the parameter value A set in the parameter setting unit 1a is 0, the process proceeds to step S14, and whether the parameter value B set in the parameter setting unit 1a is the minimum value. Judge whether.
[0040]
When it is determined in step S14 that the set parameter value B is not the minimum value, the process proceeds to step S15, and the capture control unit 3a decreases the parameter value B in the parameter setting unit 1a by a predetermined value, and the process proceeds to step S3. Return. If it is determined in step S14 that the value is the minimum value, the process proceeds to step S7, and the capture control unit 3a outputs the captured image data to the matching unit 3b.
Y_th2In the case of <Y, since the captured image is in a slightly dark state, the fingerprint values are captured again by reducing the parameter values A and B as described above.
[0041]
  The capture control unit 3a repeats the above series of operations until the image quality in the image data satisfies a predetermined condition, and when the condition is satisfied, outputs the image data as an image suitable for authentication to the matching unit 3b. in this way, SettingAn image in a desired state that falls within the specified parameter range can be obtained.
  Thereafter, the collation means 3b that has received the image data receives the received image data and the registered image data G stored in the storage unit 4.₁~ G_NAuthentication processing is performed by comparing with.
[0042]
In the above description, the density index is used as the image quality evaluation index of the captured image data. However, the present invention is not limited to this. For example, saturation, brightness, contrast, and the like may be used as indices.
Further, a spatial frequency component may be extracted from the captured image data by performing a process such as Fourier transform and used as an index. In this case, it is possible to evaluate whether the response of the fingerprint is appropriately acquired as an image as well as the shading. In addition, the image may be evaluated by combining this with a density index.
[0043]
  WhenHowever, in the above-described embodiment, the gray scale index is used as the image quality evaluation index. However, since the thickness and interval of the ridges of the fingerprint pattern differ depending on the target person, the image having the image quality optimized by the gray scale index. In some cases, the data cannot be said to be image data having an image quality optimal for authentication. In the following embodiment, for example, a histogram index is used for evaluation of image quality, and the above is solved. The ridge is a line indicating a pattern obtained by the convex portion of the fingerprint.
[0044]
As shown in FIG. 7, it is possible to calculate a histogram index generated from the captured image data from a histogram representing the density of the image indicated by the image data. In FIG. 7, the horizontal axis indicates the gradation value, and the vertical axis indicates the frequency. The gradation value indicates the degree of shade of black and white in 256 levels (gradation), where gradation 0 is black and gradation 255 is white. From the captured image data as shown in FIG. 7A, a histogram index is calculated as shown in FIG. 7B, and from the captured image data as shown in FIG. A histogram index is obtained as shown in FIG.
[0045]
By the way, normally, when image data is captured as shown in FIG. 8A, two convex portions indicating the density of ridges and noise appear in the histogram as shown in FIG. 8B. . However, as will be described later, a ridge is provided as shown in FIGS. 7B and 7D by providing a calibration circuit in the image capturing unit 1 (FIG. 1) to suppress and reduce noise and the like. It is also possible to cause only one convex portion showing the density of the color to appear in the histogram.
[0046]
In the present embodiment, the peak value is the frequency of the maximum value on the low gradation level indicated by the histogram index, and the gradation side where the gradation of the image is dark, and the image value is higher than the gradation value indicating the peak value. The frequency of the local minimum value closest to the peak value on the light gray level side is defined as the tail value, and the histogram index H is defined as H = peak value / tail value. In this embodiment, since the gradation value 255 is white, the tail value is present on the right side of the peak value in the histogram of FIG. In FIG. 8, the convex portion indicating the density of the ridge exists on the gradation side where the gradation value of the image is dark in the histogram. Therefore, the minimum value between the two convex portions is the tail value. The maximum value of the convex part on the left side of is the peak value.
[0047]
The histogram index H defined as described above has a small image contrast when the value is small, and the contrast is large when the value is large. In the present embodiment, parameters A and B are set based on such a histogram index H. In the case of black and white unclear image data as shown in FIG. 7A, the histogram index H (= peak value / tail value) is small as shown in FIG. 7B.
[0048]
In such a case, the values of the parameters A and B are set so that the histogram index H becomes large so that a result as shown in FIG. As a result, the contrast of the fingerprint image is increased as shown in FIG. 7C, and a clear fingerprint image can be obtained. In this way, by setting the parameters A and B so that the histogram index H is within a preset reference value range, that is, an appropriate range, fingerprint authentication is performed without being affected by the density of the fingerprint. An appropriate fingerprint image can be acquired.
[0049]
Next, the operation of the fingerprint collation apparatus (image capturing apparatus) using a histogram index as an evaluation index will be described using the flowchart of FIG.
First, the capture control unit 3a of the control unit 3 instructs the user who performs collation from the finger placement instruction unit 6 to place the finger on the detection surface 12 (FIG. 2) of the image capture unit 1 (step S1). As specific examples of the finger placement instruction, there are a form in which an instruction is displayed using a display device using a light emitting element such as an LED, and a form in which a predetermined instruction ON, a voice message, or the like is output. After the finger placement detection unit 7 detects that the user has placed the finger on the detection surface 12 according to the finger placement instruction (step S2), the image capture unit 1 captures the fingerprint image and outputs the image data 2 (step S3). ). These are the same as in the embodiment shown in FIG.
[0050]
In the present embodiment, when the image data 2 is output, the capture control unit 3a calculates a histogram index H from the image data 2 as an image quality evaluation index of the captured image (step S101). The histogram index H may be calculated not only from all the pixels of the captured image but also from a predetermined region of the captured image, for example, a region provided near the center of the detection surface 12. In this case, the histogram index H can be calculated in a shorter time than when all pixels are targeted.
[0051]
In the next step S102 to step S112, the capture control unit 3a determines the histogram index H obtained as described above and the preset threshold value H._th1And H_th2And compare.
First, the capture control unit 3a determines that the calculated histogram index H is the threshold value H in step S102._th1Judge whether it is above. In step S102, the histogram index H is a threshold value H._th1If it is determined as above, the process proceeds to step S103, and this time, the histogram index H is set to the threshold value H._th2Determine if: In step S103, the histogram index H is changed to the threshold value H._th2If it is determined as follows, the process proceeds to step S104, and the capture control unit 3a outputs the captured image data to the matching unit 3b.
[0052]
On the other hand, in step S102, the histogram index H is the threshold value H._th1If it is determined that the value is less than the value, the process proceeds to step S105, and first, it is determined whether or not the parameter value B set in the parameter setting unit 1a is the maximum value.
If it is determined in step S105 that the set parameter value B is not the maximum value, the process proceeds to step S106, and the capture control unit 3a increases the parameter value B in the parameter setting unit 1a by a predetermined value, and then proceeds to step S3. Return. If it is determined in step S105 that the parameter value B set in the parameter setting unit 1a is the maximum value, the process proceeds to step S107, and the parameter value A set in the parameter setting unit 1a is the maximum value. Determine whether or not.
[0053]
If it is determined in step S107 that the set parameter value A is not the maximum value, the process proceeds to step S108, and the capture control unit 3a increases the parameter value A in the parameter setting unit 1a by a predetermined value, and then proceeds to step S3. Return. If it is determined in step S107 that the value is the maximum value, the process proceeds to step S104, and the capture control unit 3a outputs the captured image data to the matching unit 3b.
H <H_th1In this case, since the contrast of the captured image is slightly small, the fingerprint image is captured again by increasing the parameter values A and B as described above.
[0054]
In step S103, the histogram index H is set to the threshold value H._th2If it is determined that the value exceeds the value, the process proceeds to step S109, where it is determined whether or not the parameter value A set in the parameter setting unit 1a is a minimum value, for example, zero.
If it is determined that the set parameter value A is not 0, the process proceeds to step S110, the capture control unit 3a decreases the parameter value A in the parameter setting unit 1a by a predetermined value, and the process returns to step S3. If it is determined in step S109 that the parameter value A set in the parameter setting unit 1a is 0, the process proceeds to step S111, and whether the parameter value B set in the parameter setting unit 1a is the minimum value. Judge whether.
[0055]
If it is determined in step S111 that the set parameter value B is not the minimum value, the process proceeds to step S112, and the capture control unit 3a decreases the parameter value B in the parameter setting unit 1a by a predetermined value, and the process proceeds to step S3. Return. If it is determined in step S111 that the value is the minimum value, the process proceeds to step S104, and the capture control unit 3a outputs the captured image data to the matching unit 3b.
H_th2In the case of <H, since the contrast of the captured image is slightly high, the fingerprint values are captured again by reducing the parameter values A and B as described above.
[0056]
The capture control unit 3a repeats the above series of operations until the image quality in the image data satisfies a predetermined condition, and when the condition is satisfied, outputs the image data as an image suitable for authentication to the matching unit 3b. Thus, according to the present embodiment, it is possible to acquire a fingerprint image with clear contrast without being affected by the density of the fingerprint pattern, individual differences, and state changes.
[0057]
As an exception process, if the condition is not satisfied even when the values of the parameters A and B reach the maximum value or the minimum value, the captured image at this time is output. Thereafter, the collation means 3b that has received the image data receives the received image data and the registered image data G stored in the storage unit 4.₁~ G_NAuthentication processing is performed by comparing with.
In the present embodiment, the above operation can also be performed after smoothing the histogram of the captured fingerprint image. In this way, it is possible to reduce the influence of histogram fluctuations due to noise, variations, and the like.
[0058]
Hereinafter, a configuration example of the fingerprint collation apparatus (image capturing apparatus) in the present embodiment, which has been described with reference to the flowchart of FIG. 9, will be described. Hereinafter, a configuration in which a calibration circuit is provided in the image capturing unit 1 (FIG. 1) in order to suppress and reduce noise and the like will be described. FIG. 10 is a configuration diagram illustrating a configuration of the image capturing unit 1 provided with the calibration circuit 101. In FIG. 10, a single sensor cell is shown as a representative in consideration of the fact that many sensor cells have the same configuration.
[0059]
As shown in FIG. 10, the sensor cell 11 includes a detection element 11a, a sensor circuit 11b, and an output signal level correction circuit 101. The output signal level correction circuit 101 takes in the output of the sensor circuit 11b, and the sensor circuit 11b The output is supplied to one of the inputs of the comparison circuit 102. The comparison circuit 102 compares the output of the sensor circuit 11 b with the calibration reference value (signal) that is the output of the calibration reference signal generation circuit 103. The calibration circuit 105 adjusts (corrects) the output level of the sensor circuit 11b by gain control of the input side of the sensor circuit 11b or the sensor circuit 11b so that the difference between the two output from the comparison circuit 102 is eliminated.
[0060]
The calibration reference signal generation circuit 103 is preferably configured to generate a calibration reference signal at the same level for each sensor cell.
The output signal level correction circuit 101 has a function of acting as much as possible to eliminate variations in the output of each sensor cell 11, and various known configurations can be applied as long as they operate in this way. In this way, the output of each sensor circuit, that is, the output level of each sensor cell 11 can be adjusted to be the same, noise due to sensitivity variations can be suppressed, and a peak due to noise in the histogram index as shown in FIG. Can be eliminated.
[0061]
  This will be described in more detail below. FIG. 11 is a configuration diagram illustrating a partial configuration of the image capturing unit 1 by the sensor cell 11 including the calibration circuit 105. Each sensor cell 11 has the same configuration, and includes a detection element 11a, a sensor circuit 11b, and a calibration circuit (sensitivity adjustment circuit) 105, and adjusts the detection sensitivity of each sensor cell 11 using the calibration circuit 105. . Calibration circuit 105, signal processing circuit110And control line L_CIt has.
[0062]
  The sensor cell 11 includes a detection element 11a, a sensor circuit 11b, and a calibration circuit 105. The detection element 11a is an element for converting the surface shape into an electrical signal. The sensor circuit 11b is a circuit that measures the amount of electricity of the detection element 11a that varies depending on the surface shape.
  When the output level of each sensor cell 11 is corrected, i.e., calibration is performed, a reference sample without unevenness is detected by the sensor cell 11 as an object to be measured, or detection is performed without placing anything on the sensor cell 11. Detect the same measured value. The signal output from the sensor cell 11 is the data line L_DA / D converter circuit via14 is output as a digital output signal 4A.
[0063]
  A / D conversion circuit1The digital output signal 4A output from 4 is a signal processing circuit.110Is also entered. Signal processing circuit110A / D conversion circuit14 is compared with a digital output signal to be originally output (hereinafter referred to as an expected value) to calculate an adjustment parameter for adjusting the detection sensitivity of the sensor circuit 11b. Then, based on the calculated adjustment parameter, the control line L_CIs used to control the calibration circuit 105.
[0064]
  Data line L_DAnd control line L_CAre shared by the sensor cells 11, and the sensor cells 11 are sequentially selected, and the output signal 2A of the sensor cells 11 is sequentially converted into an A / D conversion circuit.14 and signal processing circuit110Thus, the calibration circuit 105 in the sensor cell 11 is controlled.
  The sensitivity of each sensor circuit 11b is adjusted by repeating this operation once or a plurality of times for each sensor cell 11, and the performance of each sensor cell 11 is made uniform.
[0065]
  In this case, the signal processing circuit110Includes the comparison circuit 102 described in FIG. 10 and the calibration reference signal generation circuit 103 as another signal processing circuit 110. In the example of FIG. 11, the input signal is digital, and when the digital signal is input to the comparison circuit 102 as it is, a known digital comparison circuit can be used as the comparison circuit 104. If the comparison circuit 104 is a normal analog comparison circuit, it is once D / A converted and supplied to the comparison circuit 104. Of course, the same applies to the calibration reference signal generation circuit 103.
[0066]
In addition, as shown in FIG. 12, an image capturing unit 1 using a sensor cell 11 including a calibration circuit 105 may be configured. The image capturing unit 1 shown in FIG. 12 uses a sensor circuit 11 b ′ with a voltage-time conversion function and a time signal comparison circuit 106. The sensor circuit 11b 'with a voltage-time conversion function is a sensor circuit that converts an output signal corresponding to the amount of electricity from the corresponding detection element 11a into a signal that changes in the time axis direction. The time signal comparison circuit 106 compares the voltage-time conversion signal output from the sensor circuit 11b ′ with the voltage-time conversion function and the calibration reference signal, and uses the signal difference as a control signal for the calibration circuit 105. Output to. When the calibration of each sensor cell 11 of the image capturing unit 1 configured as described above is performed, a reference sample without unevenness is detected by the sensor cell 11 as an object to be measured, and the same measurement value is detected by each sensor cell 11.
[0067]
As a result, the sensor circuit 11b ′ with voltage-time conversion function converts a signal having analog information as a voltage value into a signal having analog information in the time axis direction as an output signal 2B as shown in FIG. Output (see FIG. 14: t_SIs the output time, and this t_SChanges). The output signal 2B is supplied from the data line L_DIs input to the A / D conversion circuit 4 via the signal and output as a digital output signal. At the same time, the output signal 2B is supplied to the calibration circuit 105 via the time signal comparison circuit 106 inside the sensor cell 11. The time signal comparison circuit 106 corresponds to the comparison circuit 102 of FIG. 10, and the digital output 2B of the sensor circuit 11b 'with the voltage-time conversion function and the reference time t from the calibration reference signal generation circuit._RThe time difference from the reference pulse signal having
[0068]
The time signal comparison circuit 106 compares the signal voltage-time converted by the sensor circuit 11 b ′ with the voltage-time conversion function with a reference pulse signal, and sends a comparison pulse signal representing the time difference to the calibration circuit 105. As a result, the calibration circuit 105 performs a control operation so that there is no time difference between the reference pulse signal and the output from the sensor circuit 11b 'with the sensor circuit voltage-time conversion function. Data line L_DAre shared by the plurality of sensor cells 11, and the sensor cells 11 are selected in sequence, and the above operation is performed.
[0069]
By repeating the above operation once or a plurality of times for each sensor cell 11, the sensitivity of each sensor circuit 11b 'with voltage-time conversion function is adjusted, and the performance of each sensor cell 11 is made uniform. Further, even if a plurality of sensor cells 11 are selected at the same time, calibration can be performed in parallel for each sensor cell, and calibration can be performed at high speed.
[0070]
FIG. 13 shows a specific configuration example of a sensor cell having a sensor circuit with a voltage-time conversion function. The sensor circuit 11b ′ with a voltage-time conversion function includes, for example, a voltage-time conversion circuit 121, and a general-purpose circuit may be used as the voltage-time conversion circuit 121. You may comprise from the threshold circuit 121a (for example, refer Japanese Patent Application No. 11-157755).
[0071]
The operation principle of the calibration circuit when using the sensor circuit 11b 'with the voltage-time conversion function will be briefly described with reference to FIG.
First, the time signal comparison circuit 106 is typically constituted by an AND circuit, and outputs an output signal 2B of the sensor circuit 11b ′ with voltage-time conversion function and a reference pulse signal from a reference pulse signal generation circuit (not shown). The logical product is taken, and the result is output to the counter circuit 154 as a comparison pulse signal.
[0072]
The value of the counter circuit 154 is the value of all load elements Z constituting the load circuit 151.₁~ Z_NIs set in advance to an inactive state, and the output signal 2B of the sensor circuit 11b 'with the voltage-time conversion function is also set to the initial set value. In the first sense operation, when the output signal 2B changes earlier than a predetermined time (for example, the sense operation is performed without placing an object such as the finger 21 on the detection surface 12 of the image capturing unit 1, and during the sense time) When the output signal of the sensor cell changes), the counter circuit 154 counts up by one based on the output of the time signal comparison circuit 106. As a result, the data of the counter circuit 154 becomes the load element Z₁~ Z_NChanges to activate one. Further, the output signal 2B of the sensor circuit 11b 'with the voltage-time conversion function is also returned to the initial set value.
[0073]
In the second sensing operation, when the output signal changes earlier than a predetermined time (for example, when the sensing operation is performed without placing the finger 13 and the output signal of the sensor cell changes during the sensing time), the counter circuit 154 Is incremented by one. Where load element Z₁~ Z_NIf the value of is set to double in accordance with the digit of the counter, as a result, twice as many load elements are activated. For example, Z₁= Z, Z₂= 2Z, Z_Three= 4Z, ..., Z_N= 2^(N-1)Z is set in advance, and Z is sequentially from the lower bit of the counter circuit 154.₁~ Z_NThe load element Z connected to the sensor circuit 11b 'with voltage-time conversion function.₁~ Z_NThe value of Z increases by Z every count-up.
[0074]
This operation is performed until the output signal of the sensor circuit 11b ′ with voltage-time conversion function does not change during a predetermined time (for example, the sensing operation is performed without placing the finger 13 and the output signal of the sensor cell is detected during the sensing time). Repeat until is no longer changing. When the output signal does not change during the predetermined time, the counter circuit 154 is not counted up, and the load element is no longer connected to the sensor circuit 11b 'with voltage-time conversion function.
[0075]
As described above, since voltage-time conversion causes no direct current to flow in the output level correction system including calibration, the power consumption of the entire apparatus can be reduced as compared with the other embodiments.
As described above, the calibration circuit 105 is added to the sensor circuit 11b, and the load element Z inside the calibration circuit 105 is added.₁~ Z_NBy connecting an appropriate number of sensors to the sensor circuit 11b, the sensor cell performance variations due to process variations and the like become invisible, and as a result, the performance of each sensor cell can be made uniform.
[0076]
<Embodiment 3>
Next, an image capturing method according to another embodiment of the present invention will be described. In the present embodiment, the ridge number index is used as the image quality evaluation index. First, the ridge number index will be described with reference to FIG. Fingerprint ridges appear as low gradation (black) lines in a fingerprint image. In the fingerprint determination area (n × m pixels) shown in FIG. 15, the ridge count index N counts the average number of ridges of the fingerprints in the horizontal and vertical directions. The line number index N is used.
[0077]
How to count the number of ridges is explained. For example, when counting the number of ridges in the horizontal direction, pixels are scanned in the vertical direction, and the ridges crossing at this time are counted. Assuming that the number of ridges when scanning one column of pixels in the vertical direction is Ni, since there are n determination areas, the average number of ridges for n columns is expressed by the following equation (1).
[0078]
[Expression 1]

[0079]
Since the length of one column in the vertical direction is m pixels, by multiplying Eq. 1 by P / m to normalize the number of ridges per P pixel, the number of pixels P is expressed as shown in Eq. 2 below. The average number of ridges Nv in the horizontal direction is calculated.
[0080]
[Expression 2]

[0081]
When the vertical direction is determined in the same manner, the average number of ridges Nh in the vertical direction per pixel number P is expressed by the following formula 3.
[0082]
[Equation 3]

[0083]
Therefore, the ridge number index N is expressed as N = MAX (Nv, Nh), and the larger of Nv and Nh is defined as the ridge number index N.
[0084]
FIG. 16 is a diagram for explaining ridge determination conditions. In FIG. 16, the vertical axis represents the gradation value of each pixel, and the horizontal axis represents the pixel in the vertical or horizontal direction of the determination area. The gradation value of each pixel is wavy due to the unevenness of the fingerprint image, and the number of waves is the number of ridges. For example, the number of ridges is counted by using two threshold values HVTH and LVTH, and counting one ridge when LVTH crosses LVTH. Here, the two threshold values are used in order to prevent pixels due to noise or the like from being counted as ridges, but of course, one threshold value can be set as HVTH = LVTH.
[0085]
It is also possible to count the ridges after smoothing the gradation value of the fingerprint image. In this way, it is possible to reduce the influence of fluctuations in the fingerprint image due to noise, device variations, and the like.
[0086]
When the value of the ridge number index N is extremely small, the fingerprint image is crushed. On the other hand, when the value of the ridge number index N is large, the fingerprint image is blurred and the ridges are broken. Accordingly, when the ridge number index N is extremely small, the image quality is thinned so that the image is not crushed. Conversely, when the ridge number index N is extremely large, the image is darkened. , Parameters A and B are adjusted. By doing so, it is possible to acquire an appropriate fingerprint image that is not affected by the density of the fingerprint pattern.
[0087]
Note that the ridge number index can be calculated from the valley line (white line in the fingerprint image of FIG. 15) instead of the ridge line (black line in the fingerprint image of FIG. 15). In this case, the evaluation index is equivalent to the ridge number index. For example, FIG. 16 explains how to count the valley lines. When the two thresholds HVTH and LVTH cross LVTH to HVTH, one valley line can be counted. Also in this case, the threshold value can be made one.
[0088]
Also in this embodiment, the configuration of the apparatus is the same as that of the image capturing apparatus (fingerprint collating apparatus) similar to the other embodiments described above, and the capturing control unit 3a operates as shown in FIG. is there. Hereinafter, the operation of the capture control unit 3a in the fingerprint collation apparatus according to the present embodiment will be described. Note that description of structural requirements equivalent to those in FIG. 9 will be omitted as appropriate.
[0089]
First, the finger placement instruction unit 6 instructs the user performing collation to place the finger on the detection surface 12 (FIG. 2) of the image capturing unit 1 (step S1). As specific examples of the finger placement instruction, there are a form in which an instruction is displayed using a display device using a light emitting element such as an LED, and a form in which a predetermined instruction ON, a voice message, or the like is output. After the finger placement detection unit 7 detects that the user has placed the finger on the detection surface 12 according to the finger placement instruction (step S2), the image capture unit 1 captures the fingerprint image and outputs the image data 2 (step S3). ).
[0090]
When the image data 2 is output, the capture control unit 3a first calculates the ridge number index N of the captured image as an image quality evaluation of the image data 2 (step S201). The ridge number index N may be calculated not only from all the pixels of the captured image but also from a predetermined region of the captured image, for example, a region provided near the center of the detection surface 12. In this case, the ridge number index N can be calculated in a shorter time than when all pixels are targeted.
[0091]
In the next step S202 to step S212, the capture control unit 3a determines the ridge number index N obtained as described above and a preset threshold value N._th1And N_th2And compare.
First, the capture control unit 3a determines that the calculated ridge number index N is N in step S202._th2If it is determined as follows, the process proceeds to step S203, and this time, the ridge number index N is a threshold value N_th1Judge whether it is above. In step S203, the ridge number index N is set to the threshold value N._th1If it is determined as described above, the process proceeds to step S204, and the capture control unit 3a outputs the captured image data to the collating unit 3b.
[0092]
On the other hand, in step S202, the ridge number index N is a threshold value N._th2If it is determined that the parameter value B is exceeded, the process proceeds to step S205, and first, it is determined whether or not the parameter value B set in the parameter setting unit 1a is the maximum value.
If it is determined in step S205 that the set parameter value B is not the maximum value, the process proceeds to step S206, and the capture control unit 3a increases the parameter value B in the parameter setting unit 1a by a predetermined value, and then proceeds to step S3. Return. If it is determined in step S205 that the parameter value B set in the parameter setting unit 1a is the maximum value, the process proceeds to step S207, and the parameter value A set in the parameter setting unit 1a is the maximum value. Determine whether or not.
[0093]
If it is determined in step S207 that the set parameter value A is not the maximum value, the process proceeds to step S208, and the capture control unit 3a increases the parameter value A in the parameter setting unit 1a by a predetermined value, and then proceeds to step S3. Return. If it is determined in step S207 that the value is the maximum value, the process advances to step S204, and the capture control unit 3a outputs the captured image data to the matching unit 3b.
N> N_th2In this case, since the captured image is too thin and too thin, the fingerprint image is captured again by increasing the parameter values A and B as described above.
[0094]
It is determined in step S203 that the ridge number index N is a threshold value N_th1When it is determined that the value is less than the value, the process proceeds to step S209, where it is determined whether or not the parameter value A set in the parameter setting unit 1a is a minimum value, for example, 0.
If it is determined that the set parameter value A is not 0, the process proceeds to step S210, and the capture control unit 3a decreases the parameter value A in the parameter setting unit 1a by a predetermined value, and returns to step S3. If it is determined in step S209 that the parameter value A set in the parameter setting unit 1a is 0, the process proceeds to step S211 to determine whether the parameter value B set in the parameter setting unit 1a is the minimum value. Judge whether.
[0095]
If it is determined in step S211 that the set parameter value B is not the minimum value, the process proceeds to step S212, and the capture control unit 3a decreases the parameter value B in the parameter setting unit 1a by a predetermined value, and the process proceeds to step S3. Return. If it is determined in step S211 that the value is the minimum value, the process proceeds to step S204, and the capture control unit 3a outputs the captured image data to the matching unit 3b.
N <N_th1In this case, since the captured image is in a crushed state, the fingerprint values are captured again by reducing the parameter values A and B as described above.
[0096]
The capture control unit 3a repeats the above series of operations until the image quality in the image data satisfies a predetermined condition, and when the condition is satisfied, outputs the image data as an image suitable for authentication to the matching unit 3b. Thus, according to the present embodiment, it is possible to acquire a fingerprint image that is not crushed or rubbed without being affected by the density of the fingerprint pattern, individual differences, and state changes.
As an exception process, if the condition is not satisfied even when the values of the parameters A and B reach the maximum value or the minimum value, the captured image at this time is output.
In the present embodiment, the location and size of the fingerprint determination area can be freely set as necessary.
[0097]
<Embodiment 4>
In this embodiment, a histogram index and a ridge number index are combined and applied to evaluation indexes in the image processing apparatus and the fingerprint collation apparatus.
For example, the parameters A and B may be changed so that each of the histogram index and the ridge number index is within a predetermined range.
Further, a new index may be derived from the histogram index and the ridge number index by a predetermined calculation, and the parameters A and B may be changed based on this index.
[0098]
By adopting such a configuration, it is possible to obtain a fingerprint image that has a strong contrast and is not crushed or rubbed without being affected by the density of the fingerprint pattern, individual differences, and state changes.
In the present embodiment, the ridge number index may be replaced with the ridge number index calculated using the valley line described in the second embodiment.
[0099]
In the above-described embodiment, the case where the analog value is converted into the digital value of 256 gradations has been described. However, the conversion to a gradation other than 256 is also possible.
In the present invention, the parameter value A and the parameter value B are used as parameters set in the parameter setting unit 1a and used for conversion in the image capturing unit 1. However, the present invention is not limited to this. For example, the same applies to the case where the minimum value and maximum value of the signal conversion region are used as the parameter value A and the parameter value B, respectively, in order to control the luminance and resolution. Further, the number of parameters to be set is not limited to this, and can be freely changed as necessary.
[0100]
In the above embodiment, as an evaluation index, a histogram generated from a combination of a density index indicating the degree of density of image data and an index indicating the spatial frequency component of the image, and a histogram expressing the density of the image. Although the combination of the index and the ridge number index generated based on the number of ridges in the image has been illustrated, it is not limited to these combinations. Needless to say, the evaluation indexes may be appropriately combined and used as new evaluation indexes.
[0101]
【The invention's effect】
As described above, according to the present invention, for example, image data in which an image evaluation index falls within a certain range is captured as a target for fingerprint collation, so that an image in a desired state can be obtained. As a result, for example, fingerprint collation accuracy can be improved.
[Brief description of the drawings]
FIG. 1 is a configuration diagram illustrating a configuration of a fingerprint collation device using an image capturing device according to an embodiment of the present invention.
FIG. 2 is a perspective view showing a partial configuration of an image capturing unit of the fingerprint collation device of FIG. 1;
FIG. 3 is an explanatory diagram illustrating a state of image data of a fingerprint captured by an image capturing unit.
FIG. 4 is a configuration diagram illustrating a schematic configuration of an image capturing unit.
FIG. 5 is an explanatory diagram showing an example of parameters set in the A / D conversion circuit 14;
FIG. 6 is a flowchart showing an operation of the fingerprint collation device according to the embodiment of the present invention.
FIG. 7 is an explanatory diagram showing captured image data and a histogram expressing the lightness and darkness of the image indicated by the image data.
FIG. 8 is an explanatory diagram showing captured image data and a histogram expressing the lightness and darkness of the image indicated by the image data.
FIG. 9 is a flowchart showing the operation of the fingerprint collation apparatus according to another embodiment of the present invention.
FIG. 10 is a configuration diagram showing a partial configuration of a fingerprint collation apparatus according to another embodiment of the present invention.
FIG. 11 is a block diagram showing a partial configuration of a fingerprint collation apparatus according to another embodiment of the present invention.
FIG. 12 is a block diagram showing a partial configuration of a fingerprint collation apparatus according to another embodiment of the present invention.
FIG. 13 is a circuit diagram showing a partial configuration of a fingerprint collation apparatus according to another embodiment of the present invention.
14 is a timing chart showing a specific operation example of the fingerprint collation device shown in FIG.
FIG. 15 is an explanatory diagram for explaining a ridge index;
FIG. 16 is a characteristic diagram for explaining ridge determination conditions;
FIG. 17 is a flowchart showing the operation of the fingerprint collation device according to another embodiment of the present invention.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Image acquisition part, 1a ... Parameter setting part, 2 ... Image data, 3 ... Control part, 3a ... Acquisition control part, 3b ... Collation means, 4 ... Memory | storage part, 5 ... Collation result, 6 ... Finger placement instruction | indication part, 7: Finger placement detection unit, G₁~ G_N... Registered image data.

Claims (11)

An image capture unit for outputting image data representing an image into an electrical quantity in accordance with the parameter value set the shape of the fingerprint to the parameter setting unit, corresponding to the shape of the fingerprint,
The image data output from the image capturing unit is received, an evaluation index for evaluating the image quality of the image is calculated from the image data, and if the evaluation index is out of a preset reference value range, Capture control for changing the parameter value set in the parameter setting unit so as to be within the range of the reference value, and outputting the image data received from the image capture unit whose evaluation index is within the range of the reference value and a part,
The image capturing unit
A detection element for converting the shape of the fingerprint into an analog signal;
An analog signal output from the detection element is composed of an A / D conversion circuit that converts the analog signal into a digital signal according to the parameter value set in the parameter setting unit and outputs the digital signal, and
The parameter value set in the parameter setting unit is a conversion range and conversion resolution when converting an analog signal into the digital signal,
The evaluation index calculated by the capture control unit is a histogram index generated from a histogram representing the density of the image,
The capture control unit
In the histogram, the histogram index is a ratio between the maximum value on the gradation side where the density of the image is dark and the minimum value closest to the maximum value on the gradation side where the density of the image is lighter than the gradation value indicating the maximum value. An image capturing device characterized by being calculated as: An image capturing unit that converts a fingerprint shape into an electrical quantity according to a parameter value set in a parameter setting unit, and outputs image data representing an image corresponding to the fingerprint shape;
The image data output from the image capturing unit is received, an evaluation index for evaluating the image quality of the image is calculated from the image data, and if the evaluation index is outside a preset reference value range, Capture control for changing the parameter value set in the parameter setting unit so as to be within the range of the reference value, and outputting the image data received from the image capture unit whose evaluation index is within the range of the reference value Department and
With
The image capturing unit
A detection element for converting the fingerprint shape into an analog signal;
An analog signal output from the detection element is composed of an A / D conversion circuit that converts the analog signal into a digital signal according to the parameter value set in the parameter setting unit and outputs the digital signal, and
The evaluation index calculated by the capture control unit is a ridge number index generated based on the number of ridges in the image,
The capture control unit
The average number of horizontal ridges, which is the average number of ridges per unit length in the horizontal direction of the image, and the number of vertical ridges, which is the average number of ridges per unit length, in the vertical direction of the image. ,
The larger of the horizontal average ridge count and the vertical ridge count is calculated as the ridge count index.
An image capturing device characterized by that. The image capturing device according to claim 1 or 2,
The image capturing device, wherein when the parameter value set in the parameter setting unit is changed , the image capturing unit outputs image data converted again according to the changed parameter value . Generating image data representing an image corresponding to the fingerprint shape by converting the fingerprint shape into an electrical signal in accordance with a preset parameter value;
From this image data, an evaluation index for evaluating the image quality of the image is calculated,
The parameter value is changed so that the evaluation index is within a preset reference value range.
An image capture method,
The parameter value is a conversion range and conversion resolution when converting an analog signal into a digital signal,
Converting the fingerprint shape into an analog signal;
This analog signal is converted into a digital signal according to the parameter value and output as the image data,
The evaluation index is a histogram index generated from a histogram expressing the density of the image,
This histogram index is a ratio between the maximum value on the gradation side where the image is dark in the histogram and the minimum value closest to the maximum value on the gradation side where the image density is lighter than the gradation value indicating this maximum value. Is
An image capturing method characterized by the above . Generating image data representing an image corresponding to the fingerprint shape by converting the fingerprint shape into an electrical signal in accordance with a preset parameter value;
From this image data, an evaluation index for evaluating the image quality of the image is calculated,
The parameter value is changed so that the evaluation index is within a preset reference value range.
An image capture method,
The parameter value is a conversion range and conversion resolution when converting an analog signal into a digital signal,
Converting the fingerprint shape into an analog signal;
The evaluation index is a ridge number index generated based on the number of ridges in the image,
The ridge number index is the larger of the average number of ridges per unit length in the horizontal direction and the average number of ridges per unit length in the vertical direction of the image.
An image capturing method characterized by the above . An image capturing unit that converts unevenness due to a fingerprint into an electric quantity according to a parameter value set in a parameter setting unit, and outputs image data representing a fingerprint image corresponding to the unevenness of the fingerprint;
The image data output from the image capturing unit is received, an evaluation index for evaluating the image quality of the fingerprint image is calculated from the image data, and if the evaluation index is outside the preset reference value range, Capture that changes the parameter value set in the parameter setting unit so as to be within the range of the reference value, and outputs the image data received from the image capture unit whose evaluation index is within the range of the reference value A control unit;
Collation means for comparing and collating image data output from the capture control unit with registered image data prepared in advance;
With
The image capturing unit
A detection element that converts the irregularities caused by fingerprints into analog signals;
An analog signal output from the detection element is composed of an A / D conversion circuit that converts the analog signal into a digital signal according to the parameter value set in the parameter setting unit and outputs the digital signal, and
The parameter value set in the parameter setting unit is a conversion range and conversion resolution when converting an analog signal into the digital signal,
The evaluation index calculated by the capture control unit is a histogram index generated from a histogram representing the density of the image,
This histogram index is a ratio between the maximum value on the gradation side where the density of the image is dark in the histogram and the minimum value closest to the maximum value on the gradation side where the density of the image is lighter than the gradation value indicating the maximum value. Is
A fingerprint collation device characterized by the above. An image capturing unit that converts unevenness due to a fingerprint into an electric quantity according to a parameter value set in a parameter setting unit, and outputs image data representing a fingerprint image corresponding to the unevenness of the fingerprint;
The image data output from the image capturing unit is received, an evaluation index for evaluating the image quality of the fingerprint image is calculated from the image data, and if the evaluation index is outside the preset reference value range, Capture that changes the parameter value set in the parameter setting unit so as to be within the range of the reference value, and outputs the image data received from the image capture unit whose evaluation index is within the range of the reference value A control unit;
Collation means for comparing and collating image data output from the capture control unit with registered image data prepared in advance;
With
The image capturing unit
A detection element that converts the irregularities caused by fingerprints into analog signals;
An analog signal output from the detection element is composed of an A / D conversion circuit that converts the analog signal into a digital signal according to the parameter value set in the parameter setting unit and outputs the digital signal, and
The evaluation index calculated by the capture control unit is a ridge number index generated based on the number of ridges in the image,
This ridge number index calculates the average number of ridges per unit length in the horizontal and vertical directions of the image, and is the larger of the calculated average ridge numbers in the horizontal and vertical directions.
A fingerprint collation device characterized by the above. In the fingerprint collation device according to claim 6 or 7,
  When the parameter value set in the parameter setting unit is changed, the image capturing unit outputs image data converted again according to the changed parameter value. Generate image data representing a fingerprint image corresponding to the fingerprint unevenness by converting the unevenness due to the fingerprint into an electric signal according to a preset parameter value,
From this image data, an evaluation index for evaluating the image quality of the fingerprint image is calculated,
The parameter value is changed so that the evaluation index is within a preset reference value range,
The image data whose evaluation index is within the range of the reference value is compared with the registered image data prepared in advance for collation.
A fingerprint verification method,
The parameter value is a conversion range and conversion resolution when converting an analog signal into a digital signal,
Converts irregularities caused by fingerprints into analog signals,
This analog signal is converted into a digital signal according to the parameter value and output as the image data,
The evaluation index is a histogram index generated from a histogram expressing the density of the image,
This histogram index is a ratio between the maximum value on the gradation side where the density of the image is dark in the histogram and the minimum value closest to the maximum value on the gradation side where the density of the image is lighter than the gradation value indicating the maximum value. Is
A fingerprint collation method characterized by the above. Generate image data representing a fingerprint image corresponding to the fingerprint unevenness by converting the unevenness due to the fingerprint into an electric signal according to a preset parameter value,
From this image data, an evaluation index for evaluating the image quality of the fingerprint image is calculated,
The parameter value is changed so that the evaluation index is within a preset reference value range,
The image data whose evaluation index is within the range of the reference value is compared with the registered image data prepared in advance for collation.
A fingerprint verification method,
The parameter value is a conversion range and conversion resolution when converting an analog signal into a digital signal,
Converts irregularities caused by fingerprints into analog signals,
This analog signal is converted into a digital signal according to the parameter value and output as the image data,
The evaluation index is a ridge number index generated based on the number of ridges in the image,
The ridge number index is the larger of the average number of ridges per unit length in the horizontal direction of the image and the average number of ridges per unit length in the vertical direction.
A fingerprint collation method characterized by the above. In the fingerprint collation device according to any one of claims 6 to 8,
A finger placement detection unit for detecting that a finger is placed on the image capturing unit;
When the finger placement detection unit detects that a finger is placed on the image capture unit, the image capturing unit converts unevenness due to a fingerprint into an electric quantity according to a parameter value set in a parameter setting unit, and Outputs image data that represents the fingerprint image corresponding to the unevenness of the fingerprint
A fingerprint collation device characterized by the above .

Images