JP2004361333A - Device for determining state of granular specimen - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To directly display a raw data measured by scanning and a function computed data as a feature amount data to conduct absolute evaluation not affected by differences in places of origin, kinds or the like. <P>SOLUTION: A display condition on an LCD monitor 8A is confirmed visually by an inspection person to determine quality of each of determination items. The feature amount data are the raw data and the function computed data defined univalently based on a photometric data, and a character message is only a comparison result with a national average, to eliminate a difference caused by determining the quality automatically according to a level of a threshold value. The quality is determined thereby depending on the inspection person, and the quality determination of high precision is possible resultingly, by judging visually the character message for the comparison with an accurate data and the national average by the inspection person knowing well the differences in the places of origin and the kinds of grains. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

【００６０】
上記表１内の「大きさデータ」とは、整列している穀粒を格子状に分割し、この格子内の穀粒が占有する画素数を言う。第１の実施の形態では、全画像が２５０万画素（Ａ４サイズ程度）であり、１０００粒の穀粒を領域分割すると、１つの格子が２５００画素で構成され、その中の標準的な穀粒は１５００画素程度である。
【００６１】
また、上記生データに基づいて、所定の関数演算することで関数演算データを得るようにしている。
【００６２】
関数演算データには、以下の表２に示されるような種類がある。
【００６３】
【表２】

【００６４】
コントローラ１８では、ＬＣＤモニタ１８Ａに表示する表示対象データとして、上記ｄ．〜ｆ．の生データ、並びにｇ．〜ｊ．の関数演算データの一部又は全部を数値データ或いはグラフィカルな表示形態（ここでは、レーダーチャート）で表示するようになっている。
【００６５】
この表示内容に基づいて、以下の表３に示されるような判定項目に対して、検査員の良否判定が行われるようになっている。
【００６６】
【表３】

【００６７】
図４は、例えば、Ｎｏ．ｄ、Ｎｏ．ｇ〜Ｎｏ．ｊの関数演算データをレーダーチャートとして表示したものである。このレーダーチャートは、ＬＣＤモニター１８Ａの一部に表示されるようになっている。
【００６８】
図５に示される如く、ＬＣＤモニタ１８Ａの画像表示形態は、左から２／３は、実際に穀粒を撮像した画像を表示する穀粒表示領域１００であり、整列された穀粒と共に、この穀粒を１粒毎に区分けする格子状の罫線１０２が同時に表示される。なお、この穀粒表示領域１００は、独立してスクロール可能とされ、合計１０００粒（１回の計測数）の一部が穀粒表示領域１００に表示される。
【００６９】
一方、穀粒表示領域１００の右側には、特徴量データ表示領域１０４とされ、その上部が前記レーダーチャートを表示するためのグラフィック表示領域１０６、下部が数値データ表示領域１０８となっている。
【００７０】
以下に第１の実施の形態の作用を説明する。
【００７１】
（測光手順）
トレイ７２を挿入した状態で、ホッパーユニット４４の受皿５２にオペレータの手作業によって投入する。この場合、ホッパーユニット４４は、ホームポジションセンサ５０の検出によりカバー４６の投入口４８に対応しているため、オペレータは投入口４８に適量の穀粒を投入すればよい。
【００７２】
適用の穀粒が投入され、起動を促す操作がなされると、ヘッド部３４の光源３４Ｂをオンとし撮像素子アレイ３４Ａをアクティブ状態とし、スキャナモータ３６の駆動を開始する。
【００７３】
このスキャナモータ３６の駆動により、ヘッド部３４は、スキャン開始位置センサ３８よりも若干上流側のホームポジションから移動を開始する。
【００７４】
このため、移動を開始直後には、スキャン開始位置センサ３８によってその位置が検出され、タイマをスタートさせ（ｔ１）、測距位置センサ４２によってヘッド部３４を検出すると、タイマをストップ（ｔ２）させる。これにより、スキャン開始位置センサ３８から測距位置センサ４２までのヘッド部３４の移動時間Δｔ（＝ｔ２−ｔ１）を得ることができる。ここで、前記時間Δｔと予め既知のスキャン開始位置センサ３８から測距開始位置センサ４２までの距離Ｌとに基づいてヘッド部３４の移動速度ｖ_Ｓを演算する。
【００７５】
演算されたヘッド部３４の移動速度ｖ_Ｓに基づいて、ホッパーユニット４４の移動速度とロータリバルブ５４の回転線速度とを設定し、ロータリーバルブ５４を回転させながらホッパーユニット４４を副走査方向へ移動させる。
【００７６】
これにより、被検査物としての穀粒がトレイ７２上に縦横に整列して均等に散布される。
【００７７】
この状態で、ヘッド部３４による測光を実行する。すなわち、トレイ７２上の穀粒に光源３４Ｂからの光を照射し、その反射光を撮像素子アレイ３４Ａによって検出しながらヘッド部３４を移動させることで、トレイ７２の所定の領域に散布された穀粒の個々のデータを取り込む。
【００７８】
測光が終了すると、トレイ７２を排出移動させ、次の測光のスタンバイのため、ヘッド部３４、ホッパーユニット４４をそれぞれ初期位置（ホームポジション）へ戻す。
【００７９】
コントローラ１８では、取り込んだデータを処理し、ＬＣＤモニタ１８Ａへの表示を制御し、プリント指示があった場合にプリンタ２０でのプリントを制御する。
【００８０】
（ＬＣＤモニタへの表示）
ここで、第１の実施の形態では、ＬＣＤモニタ１８Ａへの表示内容を、前記ヘッド部３４の走査によって得た、測光データに基づいて一義的に決まる生データ、並びに生データを所定の関数で演算した関数演算データで構成される特徴量データとしている。
【００８１】
表示項目は、ユーザーの指定が可能であり、生データ（表１のＮｏ．ａ〜Ｎｏ．ｆ）の内のＮｏ．ｄ〜Ｎｏ．ｆ、並びに関数演算データ（表２のＮｏ．ｇ〜Ｎｏ．ｋ）の内のＮｏ．ｇ〜Ｎｏ．ｊの中から選択できる。
【００８２】
例えば、レーダーチャートとしてＮｏ．ｄ、Ｎｏ．ｇ〜Ｎｏ．ｊの５項目を選択すると、図５に示されるようなレーダーチャートが表示される。なお、このとき、同時にレーダーチャートに載らないデータは、表形式で数値で表示される。
【００８３】
さらに、図４では、グラフィック表示領域１０６にレーダーチャート上に全国平均値に基づくデータが表示されると共に、数値データ／文字メッセージ表示領域１０８にこの全国平均値との比較結果を、比較形容詞を用いた言葉による文字メッセージを表示する。
【００８４】
例えば、「全国の平均的な米より、青くてやや細く、小さい」といった文字メッセージが表示される。
【００８５】
上記ＬＣＤモニタ１８Ａへの表示状態を、検査員（ユーザー）が目視で確認することで、判定項目（表３参照）のそれぞれについて、良否の判定を行う。
【００８６】
特徴量データは、測光データから一義的に定まる生データと関数演算データであり、文字メッセージは、全国平均との比較結果（判定ではなく）のみであるため、この時点では、しきい値の取り方によって自動的に良否を判定した場合に起こり得る差異がない。
【００８７】
言い換えれば、良否の判定は、検査員（ユーザー）に依存することになるが、良否の判定基準となるしきい値が不適当な状態で、自動判定するよりも、穀粒の産地や品種の違いを熟知している検査員が正確なデータ（特徴量データ）や全国平均との比較のための文字メッセージを目視して判断する方が、結果的には精度の高い良否判別が可能となる。
【００８８】
（第２の実施の形態）
以下、本発明の第２の実施の形態について説明する。なお、この第２の実施の形態において、第１の実施の形態と同一構成部分については、同一の符号を付して、その構成の説明を省略する。
【００８９】
第２の実施の形態の特徴は、前記第１の実施の形態において、設定しなかったしきい値を、前記表示される特徴量データや文字メッセージに基づいて、適正なしきい値を設定可能にした点にある。
【００９０】
前記第１の実施の形態では、生データ又は関数演算データを穀粒の特徴量データとしてＬＣＤモニタ１８Ａへ表示するようにしたが、この第２の実施の形態では、判定項目の判定結果を表示することが特徴となっている。
【００９１】
この判定項目の良否判定を行うためには、それぞれの判定項目（表３参照）における判定パラメータに対するしきい値が必要となる。
【００９２】
第２の実施の形態では、それぞれの判定項目に対して、予めデフォルト値が設定され、記憶されている。
【００９３】
ここで、判定項目の合否判定の際にユーザーは、デフォルトモードとユーザー設定モードの何れかを選択できるようになっている。
【００９４】
デフォルトモードを選択した場合は、前記記憶したデフォルト値がしきい値として採用され、合否の判定が行われる。
【００９５】
一方、ユーザー設定モードを選択した場合は、それぞれの判定項目に対して、産地、品種毎に複数のしきい値を任意に設定することができ、このユーザーが設定したしきい値に基づいて、各判定項目の良否判定が行われる。
【００９６】
ユーザー設定モードでは、事前に前述の特徴量データを確認することが可能であり、この特徴量データを見ることで、適正なしきい値の設定が可能となっている。
【００９７】
すなわち、必要に応じてユーザーがしきい値を設定することで、経時的に複数の穀粒の産地に適したしきい値が記憶されることになり、必要であれば、その中から選択して再利用することもできる点が、従来と相違する点である。
【００９８】
本装置では、ユーザー設定において、特徴量データ及び文字メッセージを参考にして設定できるため、産地等、比較的ローカルな領域での適正なしきい値が徐々に増加し、当該産地において使い勝手のよい、かつ適正なしきい値の設定が可能となる。
【００９９】
図６には、第２の実施の形態に係るコントローラ１８における、しきい値設定並びに良否判定のための制御ブロック図が示されている。
【０１００】
測光データは、特徴量データ生成部１１０に送出され、生データ（表１参照）と関数演算データ（表２参照）が生成される。また、測光データは、画像処理部１１２を介して、各穀粒の画像をＬＣＤ制御部１１４へ送出する。
【０１０１】
特徴量データ生成部１１０で生成された特徴量データは、判定実行指示部１１６からの指示信号に基づいて、判定項目選択部１１８へ送出される。
【０１０２】
すなわち、判定項目選択部１１８では、判定項目メモリ１２１から判定項目（表３参照）を順次読出し、この判定項目の良否判定に必要な特徴量データを特徴量データ生成部１１０から読み出す。
【０１０３】
読み出された特徴量データは、判定部１２０に送出され、適用しきい値格納部１２２に格納されているしきい値（デフォルト値或いはユーザー設定値）との比較によって良否の判定が実行される。また、良否の判定結果は、ＬＣＤ表示制御部１１４へ送出され、ＬＣＤモニタ１８Ａに表示される。
【０１０４】
ここで、適用しきい値格納部１２２には、今回比較対照として適用するしきい値が選択的に格納されるようになっており、そのしきい値は、しきい値設定部１２４によって設定される。
【０１０５】
しきい値設定部１２４には、デフォルト値メモリ１２４Ａ及びユーザー設定しきい値メモリ１２４Ｂが設けられ、予め各設定項目に対応する標準的なしきい値が記憶されている。デフォルト値は、しきい値設定部１２４に接続されているしきい値選択部１２６において、デフォルト値を使用する旨の選択操作があった場合に読み出され、適用しきい値格納部１２２へ送出される。
【０１０６】
一方、しきい値選択部１２６において、ユーザー設定する旨の選択操作があった場合には、ユーザー設定しきい値メモリ１２４Ｂに記憶されているしきい値（複数記憶されている場合は、それぞれにＩＤ（識別符号）が付与されている。）が読み出され、適用しきい値格納部１２２へ送出される。
【０１０７】
ここで、ユーザー設定しきい値メモリ１２４Ｂには、しきい値作成指示操作部１２７からの指示操作により、しきい値を追加登録できる。
【０１０８】
この場合、ＬＣＤ表示制御部１１４へユーザー設定信号を送出し、前記特徴量データ及び文字メッセージの表示を指示する。
【０１０９】
これにより、ＬＣＤモニタ１８Ａには、特徴量データ及び文字メッセージが表示され（表示形態は、第１の実施の形態と同様であってもよいし、表形式の数値でもよい）、ユーザーはこの表示内容に基づいて、しきい値入力操作部１２８の操作により、しきい値を入力する。なお、ＬＣＤモニタ１８Ａには、しきい値入力画面として、図７に示すようなウィンドウ画面１３０が表示される。
【０１１０】
ユーザーによるしきい値設定が完了すると、この設定されたしきい値がＩＤと共にしきい値設定部１２４のユーザー設定しきい値メモリ１２４Ｂに記憶される。
【０１１１】
図８には、各判定項目と、その良否判定のためのしきい値の設定手順、並びに良否判定手順が示されている。
【０１１２】
良否判定は、ステップ２００から順に行い、最終的に残ったものが整粒となる。
【０１１３】
ステップ２００及びステップ２０２は、青未熟粒の良否判定であり、白さ評価データ（関数演算データＮｏ．ｇ）と、青さ評価データ（関数演算データＮｏ．ｈ）を用いて判定する。ステップ２００で青いか否かが判断され、肯定判定されるとステップ２０２へ移行して白いか否かが判断される。このステップ２０２で否定判定（白くない）されると、青未熟粒と判定され（ステップ２０４）、肯定判定（白い）と判定されると、青死米と判定される（ステップ２０６）。ステップ２００で否定判定されると、次のステップ２０８へ移行する。
【０１１４】
ステップ２０８は、乳白粒の良否判定であり、白さ評価データ（関数演算データＮｏ．ｇ）を用いて判定する。ステップ２０８で肯定判定（白い）されると、乳白粒と判定される（ステップ２１０）。ステップ２０８で否定判定されると、次のステップ２１２へ移行する。
【０１１５】
ステップ２１２は、腹白粒の良否判定であり、白画素率を用いて判定する。白画素率とは、生データＮｏ．ａ、Ｎｏ．ｂ、Ｎｏ．ｃを用いて表され、（Ｒ＞しきい値、かつＧ＞しきい値、かつＢ＞しきい値の画素数）／１粒の画素数の演算値であり、部分的に白いか否かが判断される。ステップ２１２で肯定判定（部分的に白い）されると、腹白粒と判定される（ステップ２１４）。ステップ２１２で否定判定されると、ステップ２１６へ移行する。
【０１１６】
ステップ２１６は、白死粒（米）の良否判定であり、白さ評価データ（関数演算データＮｏ．ｇ）を用いて判定する。ステップ２１６で肯定判定（白い）されると、白死米と判定される（ステップ２１８）。ステップ２１６で否定判定されると、ステップ２２０へ移行する。
【０１１７】
ステップ２２０、ステップ２２２、ステップ２２４は砕粒の良否判定であり、大きさ（生データＮｏ．ｄ）、並びに丸さデータ（関数演算データＮｏ．ｊ）を用いて判定する。ステップ２２０で肯定判定（小さい）されると、砕粒と判定される（ステップ２２６）。ステップ２２０で否定判定されると、ステップ２２２へ移行する。
【０１１８】
ステップ２２２で肯定判定（小さい）され、次いでステップ２２４で肯定判定（丸い）と判定された場合も、砕粒と判定される（ステップ２２８）。
【０１１９】
ステップ２２２及びステップ２２４で否定判定されると、ステップ２３０へ移行する。
【０１２０】
ステップ２３０は、その他被害の判定であり、赤さ評価データ（関数演算データＮｏ．ｉ）を用いて判定する。ステップ２３０で肯定判定（赤い）されると、その他被害と判定される（ステップ２３２）。ステップ２３０で否定判定されると、ステップ２３４へ移行する。
【０１２１】
ステップ２３４は、着色粒の良否判定であり、色変化画素率データ（関数演算データＮｏ．ｋ）を用いて判定する。ステップ２３４で肯定判定（部分的に赤い）されると、着色粒と判定される（ステップ２３６）。ステップ２３４で否定判定されると、ステップ２３８へ移行する。
【０１２２】
ステップ２３８は、その他未熟粒の判定であり、大きさ（生データＮｏ．ｄ）を用いて判定する。ステップ２３８で肯定判定（小さい）されると、その他未熟粒と判定される（ステップ２４０）。ステップ２３８で否定判定されると、ステップ２４２へ移行する。
【０１２３】
ステップ２４２は、その他未熟粒の判定であり、青さ評価データ（関数演算データＮｏ．ｈ）を用いて判定する。ステップ２４２で肯定判定（青い）されると、その他未熟粒と判定される（ステップ２４４）。ステップ２４２で否定判定されると、ステップ２４６へ移行する。
【０１２４】
ステップ２４６は、その他未熟粒の判定であり、赤さ評価データ（関数演算データＮｏ．ｉ）を用いて判定する。ステップ２４６で肯定判定（赤い）判定されると、その他未熟粒と判定される（ステップ２４８）。ステップ２４６で否定判定されると、ステップ２５０へ移行する。
【０１２５】
ステップ２５０は、その他未熟粒の判定であり、白さ評価データ（関数演算データＮｏ．ｇ）を用いて判定する。ステップ２５０で肯定判定（白い）判定されると、その他未熟粒と判定される（ステップ２５２）。ステップ２５０で否定判定されると、ステップ２５４へ移行する。
【０１２６】
ステップ２５４は、胴割の良否判定であり、色変化画素率データ（関数演算データＮｏ．ｋ）を用いて判定する。ステップ２５４で肯定判定（部分的色変化あり）されると、胴割と判定される（ステップ２５６）。ステップ２５４で否定判定されると、ステップ２５８へ移行し、整粒と判定される。
【０１２７】
なお、１粒毎の良否判定結果においては、ｘ−ｙ軸を用いたグラフを用い、例えば、ｘ軸に大きさ、ｙ軸に丸さをとって、その他未熟粒の良否判定をグラフィカルに表示すればよい。１粒毎の良否判定は、穀粒表示領域１００の所望の穀粒画像（或いは罫線１０２で囲まれた領域内）にポインタを会わせてクリックすることで、該当する穀粒の判定結果がグラフィック表示領域１０６に表示される。これは、穀粒１粒毎の性状を把握するのに有利な手段である。
【０１２８】
着色粒や被害粒など、少量でも判定に影響する項目に属する可能性のある粒の性状を、装置が数値的にどうとらえるかを知ることができる。
【０１２９】
このように、第２の実施の形態では、各判定項目に標準的なしきい値であるデフォルト値を１つだけ記憶しておき、このデフォルト値が不適当な場合に、特徴量データ及び文字メッセージを見ながら、ユーザー設定によってしきい値を設定するようにした。このため、産地等、ローカル地域に直結したしきい値の設定が可能であり、誤って不適当なしきい値を設定するような不具合がない。
【０１３０】
また、デフォルト値を用いた場合には、産地、品種が異なってもある程度の精度で、良否の判定が可能であるため作業性がよい。一方、ユーザー設定の場合には、ユーザーの熟練した技能や、特徴量データを確認しながらしきい値を設定するため、しきい値設定までの作業に時間を要するが、精度の高いしきい値設定が可能となる。
【０１３１】
なお、第１の実施の形態及び第２の実施の形態では、トレイ上への穀粒の散布（整列）をハンドリング部２６を用いて自動的に行ったが、このハンドリング部２６を取り去り、穀粒を整列するための縦横に整列された貫通孔を持つ整列板を用いて、手動でガラス板ランダムに載置することで、透明ガラス板３０上に穀粒を載置する構造であってもよい。
【０１３２】
【発明の効果】
以上説明した如く本発明では、走査によって計測した生データ及び関数演算データを特徴量データとして、直接表示することで、産地、品種等の違いに影響されない、絶対評価を行うことができるという優れた効果を有する。
【図面の簡単な説明】
【図１】第１の実施の形態に係るオートローダ式判別装置の概観を示す斜視図である。
【図２】第１の実施の形態に係るオートローダ式判別装置の内部構成を示す概略図である。
【図３】第１の実施の形態に係るホッパーユニットの構造を示す断面図である。
【図４】第１の実施の形態に係るレーダーチャートの一例を示す正面図である。
【図５】第１の実施の形態に係るＬＣＤモニタの表示画面を示す正面図である。
【図６】第２の実施の形態に係るしきい値設定及び良否判定のための制御ブロック図である。
【図７】第２の実施の形態に係るしきい値ユーザー設定用ウィンドウ画面を示す正面図である。
【図８】各判定項目と、その良否判定のためのしきい値の設定手順、並びに良否判定手順を示すフローチャートである。
【符号の説明】
１０ オートローダ式判別装置（粒状被検査物状態判別装置）
１２、１４ 箱体
１６ 筐体
１８ コントローラ
１８Ａ ＬＣＤモニタ（特徴量表示手段、判定結果表示手段）
２０ プリンタ
２２ 判別装置本体
２４ スキャナ部
３０ 透明ガラス板（測光面）
３２ ライン走査型撮像ユニット
３４ ヘッド部
３６ スキャナモータ
１００ 穀粒表示領域
１０２ 罫線
１０４ 特徴量データ表示領域
１０６ グラフィック表示領域（特徴量表示手段）
１０８ 数値データ／文字メッセージ表示領域（文字メッセージ表示手段）
１１０ 特徴量データ生成部
１１２ 画像処理部
１１４ ＬＣＤ表示制御部
１１６ 判定実行指示部
１１８ 判定項目選択部
１２０ 判定部
１２１ 判定項目メモリ
１２２ 適用しきい値格納部
１２４ しきい値設定部（しきい値設定手段）
１２４Ａ デフォルト値メモリ
１２４Ｂ ユーザー設定しきい値メモリ
１２６ しきい値選択部
１２７ しきい値作成指示操作部
１２８ しきい値入力操作部
１３０ ウィンドウ画面[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention is to scan while irradiating the granular test object with light in a state where the granular test object is aligned on the flat bed-shaped photometric surface, and measure the reflected light or transmitted light thereof. The present invention relates to a granular inspection object state determination device that obtains photometric data and determines the finished state of aligned granular inspection objects.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, a grade has been determined by comparing a granular test object, for example, a grain of rice or the like, with a standard value (default value) based on an inspection method of a predetermined organization, in order to manage the standard.
[0003]
For this grain discrimination, the inspector automatically judges the quality (pass / fail) instead of visually determining the sizing (specified product) / non-sized (specified product) according to a predetermined inspection rule. (A grain discriminator) has been proposed.
[0004]
Patent Literature 1 proposes that in order to automate the inspection, it is possible to detect transmitted light of the entire rice grain and obtain an accurate inspection signal without adversely affecting a detection signal of a sensor when aligning rice grains. .
[0005]
In Patent Document 1, a sample (seed) is imaged by a CCD camera while a sample is put into a supply hopper, and is conveyed along a conveying device in one line, and the color, size, and shape of the seed are obtained from the imaged result. And the like (pass / fail) are determined based on a pre-programmed acceptance / rejection criterion, and an extraction device is used to select a passed product and a defective product. In Patent Document 1, an image range of each seed is divided into a grid and displayed on a display screen.
[0006]
Patent Document 2 discloses a rice grain quality determination device capable of detecting a large number of rice grains at a time as a sample and determining the quality of the rice grains individually for each rice grain. In Patent Literature 2, rice grains placed in a flat bed-shaped window are detected by line scanning of a scanner unit, and a flat light amount in the window is obtained.
[0007]
[Patent Document 1]
Japanese Patent No. 3058940
[Patent Document 2]
Japanese Utility Model Publication No. 7-33151
[0008]
[Problems to be solved by the invention]
However, in the above-mentioned conventional screening by inspectors or the like, the boundaries between sizing and non-sizing are not absolute depending on varieties, birth years, regions, and the like, so that individual differences occur in the screening results.
[0009]
In addition, in a sorting device (discriminating device) that automatically sorts quality, a plurality of thresholds are required according to varieties and regions. For example, even if the same brown rice is used, a plurality of sizing rate values are calculated. there were.
[0010]
More specifically, when looking at rice in Hokkaido, the sizing rate given by a Hokkaido inspector is different from the sizing rate given by a Kyushu inspector (individual differences among inspectors).
[0011]
In the automatic sorting equipment, the threshold A that the Hokkaido inspector looks at the rice in the production area and the value that the inspector in Hokkaido produces is the same as the threshold A that Kyushu looks at the rice in the production area and the value that the inspector in the Kyushu produces. Threshold B must be prepared (threshold for each production area).
[0012]
For this reason, even if there are a plurality of threshold values, only rough classification can be performed, and correct selection may not be performed due to a setting error or the like.
[0013]
In view of the above facts, the present invention directly displays raw data and function operation data measured by scanning as feature data, thereby enabling a granular coating that can be subjected to an absolute evaluation without being affected by differences in production areas, varieties, and the like. It is an object to obtain an inspection object state determination device.
[0014]
[Means for Solving the Problems]
According to the first aspect of the present invention, in a state where the granular inspection object is aligned on the flat bed-shaped photometric surface, the granular inspection object is scanned while irradiating light, and reflected light or transmitted light thereof is provided. To obtain a photometric data by measuring the granular inspected object state determination device that determines the finished state in aligned granular inspected object unit, by analyzing the photometric data of the granular inspected object The obtained raw data, and a function calculation data obtained by calculating the raw data based on a predetermined function, as a feature value display means for displaying as the feature value data of the granular inspection object, and an average value of the feature value data as a reference. Character message display means for displaying the result of comparison with the average value using an adjective character message.
[0015]
According to the first aspect of the present invention, the granular test object is aligned on the photometric surface and irradiated with light. Photometric data is obtained by measuring the reflected light or transmitted light of this light.
[0016]
The characteristic amount display means displays raw data obtained by analyzing the photometric data and function operation data obtained by calculating the raw data based on a predetermined function as characteristic amount data of the granular inspection object.
[0017]
Since the feature amount data is not a comparison result based on an artificially set threshold value, for example, when the granular test object is a grain, it can be used as an absolute evaluation parameter that is not affected by a difference in a production place or a variety. it can.
[0018]
Further, together with the feature amount data, a comparison result between the average value of the feature amount data and a reference average value is expressed by an adjective character message.
[0019]
As is well known, the "form" refers to "the form of things", "the way it is", "the shape", and "to describe the form or thing of things using other words or parables". The "adjective" refers to "a word that expresses the property / state of an object by focusing on the attribute of super-temporal persistence of the object" (from Kojien).
[0020]
In the claims, "adjective" means that it is not necessary to end with "i".
[0021]
That is, in the present invention, text messages such as "blue", "white", "red", and "large" and "small" are displayed as compared with the reference average value. Since this text message expresses only the result of comparison with the reference average value, not the judgment of pass / fail, this does not directly lead to pass / fail. This can be used as a material for empirical judgment of quality. As a result, an erroneous quality judgment is not performed, and the final judgment is made by, for example, a skilled person in the area, but by referring to the feature amount data and the text message as the quality judgment material, Almost uniform quality judgment can be performed irrespective of a skilled person or a beginner.
[0022]
According to a second aspect of the present invention, in the first aspect of the present invention, the feature amount display means displays a part or all of the feature amount data composed of the raw data and the function operation data with the reference. It is characterized in that a list is displayed graphically for each granular inspection object together with the average value.
[0023]
According to the second aspect of the present invention, by displaying a part or all of the feature amount data graphically, an inspector or the like can identify the features of the granular inspection object at a glance. Radar charts are the best graphical data. It should be noted that by displaying the reference average value together with the radar chart as a guide, this can be used as a guide for the quality judgment.
[0024]
According to a third aspect of the present invention, in the first aspect or the second aspect of the invention, based on the feature amount data, a blue immature grain, a blue dead grain, a milky white grain, a belly white grain, a white dead grain, Judgment means for judging the quality of immature grains, crushed grains, damaged grains, colored grains, and cracked grains, and a threshold value as a criterion for the quality judgment, referring to the display states of the feature quantity display means and the text message display means. It is characterized by further comprising a threshold setting means which can be set as.
[0025]
According to the third aspect of the present invention, based on the feature amount data, blue immature grains, blue dead grains, milky grains, white belly grains, white dead grains, immature grains, crushed grains, damaged grains, colored grains, body cracks The quality of the grains is determined. In this determination, a threshold value is required. Conventionally, a threshold value is selected from a default value or a plurality of threshold values set by the manufacturer. However, the default value or the threshold value set by the manufacturer is, for example, one that can be used at the national level, and thus may be an inappropriate threshold value in some regions.
[0026]
Therefore, the threshold value setting means can set a threshold value as a criterion for quality determination with reference to the display states of the feature amount display means and the text message display means.
[0027]
According to a fourth aspect of the present invention, in the third aspect of the present invention, the threshold value set by the threshold value setting means can be accumulated and stored together with an identification code.
[0028]
According to the fourth aspect, the set threshold value is stored together with the identification code by the storage means. This makes it possible to make a pass / fail decision with a threshold value for a small area, and to carry out a more accurate pass / fail decision. Also, by storing the information in the storage means, the information set once can be read out and used in the future.
[0029]
According to a fifth aspect of the present invention, in the first aspect of the present invention, the raw data includes red (R) color separation data, green (G) color separation data, and blue (B) color separation data, size data of each inspection object, length data of each inspection object, and width data of each inspection object, wherein the function operation data is whiteness evaluation data , Blueness evaluation data, redness evaluation data, roundness data, and color change pixel rate data.
[0030]
According to the invention as set forth in claim 5, the raw data is obtained by performing color separation of the photometric data, red separation data, green separation data, blue separation data, and size data of each inspection object, The length data of each inspection object and the width data of each inspection object are included.
[0031]
On the other hand, the function operation data includes whiteness evaluation data necessary for finding defects (non-regulated) such as the above-mentioned blue immature grains, milky grains, white belly grains, dead white grains, crushed grains, colored grains, and cracks. , Blueness evaluation data, redness evaluation data, roundness data, and color change pixel rate data.
[0032]
In particular, a skilled inspector can very accurately determine the quality of the sized / non-sized granules by looking at the raw data and the function operation data as described above.
[0033]
In addition, if a correct threshold value is set based on a production area and a product type, it is possible to automatically determine whether or not the sizing / non-sizing is performed with high accuracy.
[0034]
BEST MODE FOR CARRYING OUT THE INVENTION
FIG. 1 shows an autoloader type discriminating apparatus (hereinafter, simply referred to as a discriminating apparatus) 10 according to the first embodiment.
[0035]
The discriminating device 10 has a pair of boxes 12 and 14 connected by hinges (not shown) and is housed in an openable and closable briefcase-type housing 16. FIG. 1 shows a use state.
[0036]
That is, the pair of boxes 12 and 14 are placed in a state where they are opened from each other by 180 °. At the time of storage, the pair of boxes 12 and 14 rotate around the hinge so that the opening ends face each other and are closed, thereby protecting the internal precision instruments and the like.
[0037]
A controller (which may be a general-purpose personal computer) 18 and a printer 20 are housed in one of the boxes 12 (the back side in FIG. 1). The controller 18 and the printer 20 are connected by a connection cable (not shown), and the data and the like processed by the controller 18 can be printed out by the printer 20. Further, the controller 18 is provided with an LCD monitor 18A.
[0038]
Further, the controller 18 has a function of controlling the operation of the discriminating device main body 22 housed in the other box 14, and the discriminating device main body 22 operates according to an instruction from the controller 18. ing. It is preferable that the controller 18 and the discrimination device main body 22 are connected by a flexible wiring cable (not shown) such as a flexible cable which is not damaged by opening and closing the boxes 12 and 14.
[0039]
FIG. 2 shows a schematic configuration of the discrimination device main body 22.
[0040]
The discriminating apparatus main body 22 includes a scanner unit 24 serving as a base, and a kernel handling unit 26 for receiving a kernel as a granular inspection object and arranging the kernel at a photometric position by the scanner unit 24.
[0041]
The scanner unit 24 includes a box-shaped casing 28 having an open upper surface except for the peripheral edge, and a transparent glass plate 30 is fitted into the opened portion.
[0042]
A tray 72 that can be inserted into and removed from an insertion port 80 on a slit provided in a casing 28 of a handling unit 26 described below is positioned above the transparent glass plate 30. Mounting table.
[0043]
A line scanning type imaging unit 32 is provided below the transparent glass plate 30.
[0044]
The line scanning type imaging unit 32 includes an image sensor array 34A in which the image sensors are arranged in an array from the near side to the far side in FIG. 2 (main scanning direction), and a cylindrical light source provided along the image sensor array 34A. The head unit 34 includes a head unit 34B. The head unit 34 moves below the transparent glass plate 30 in the left-right direction of FIG. 2 by the driving force of the scanner motor 36 (sub-scanning direction).
[0045]
Thereby, almost the entire surface of the transparent glass plate 30 can be scanned by the image sensor array 34A.
[0046]
The line scanning type imaging unit 32 is provided with a scan start position sensor 38 and a scan end position sensor 40. By detecting the head unit 34 with the scan start position sensor 38 and the scan end position sensor 40, A scanning area is set.
[0047]
In the vicinity of the scan start position sensor 38 of the head unit 34, a distance measuring position sensor 42 for detecting the head unit 34 is separately provided. The distance measurement position sensors 42 are arranged in the sub-scanning direction at a predetermined interval L between the distance measurement position sensors 42 and the scan start position sensor 38. The scan start position sensor 38 and the distance measurement position sensor 42 sequentially detect the head unit 34 immediately after the head unit 34 starts sub-scanning. That is, a time difference Δt (t2−t1) occurs between the time t1 detected by the upstream scan start position sensor 38 and the time t2 detected by the downstream distance measurement position sensor 42.
[0048]
This time difference Δt is determined by the moving speed v of the head unit 34. _S Is applied as a parameter for calculating. Note that the distance measuring position sensor 42 is not essential, and the scan end position sensor 40 is used in place of the scan end position sensor 40, and the time difference Δt May be obtained.
[0049]
The handling part 26 is provided with a cover 46 so as to cover the upper part of the casing 28 including the upper part of the transparent glass plate 30.
[0050]
A rectangular opening is provided in the top plate portion of the cover 46 to serve as a grain input port 48.
[0051]
That is, the grain, which is the inspection object to be measured by the scanner unit 24, is supplied from the input port 48.
[0052]
The home position of the hopper unit 44 is located immediately below the inlet 48. In a normal state (standby state), the grains input from the input port 48 are delivered to the hopper unit 44. A home position sensor 50 is provided at the home position of the hopper unit 44, and the position is managed by the home position sensor 50.
[0053]
As shown in FIG. 3, the hopper unit 44 includes a tray 52 having a tapered lower end for storing the grains flowing through the input port 48.
[0054]
A rotary valve 54 is attached to the lower end opening of the receiving tray 52. The rotary valve 54 has an uneven surface in the axial direction, so that grains can be stored in the concave portion 54A. In addition, depending on the accommodation state, a part may protrude from the concave portion 54A. Further, instead of the unevenness, a plurality of independent grooves (dents) may be used.
[0055]
The rotary valve 54 protrudes from the lower end opening of the tray 52, and an elastic valve seat 56 is hung from the lower end of the tray 52 so as to hide the protruding peripheral surface.
[0056]
The hopper unit 44 moves in the sub-scanning direction by the axial movement of the wire 66, and the rotary valve 54 rotates in synchronization with the movement of the wire 66. Are arranged in a line.
[0057]
By moving the head unit 34 back and forth in a state where the grains are arranged on the tray 72, the grains on the tray 72 can be measured (imaged). The photometric data is sent to the controller 18.
[0058]
Here, the controller of the first embodiment obtains data shown in Table 1 below as raw data by analyzing photometric data.
[0059]
[Table 1]

[0060]
The “size data” in Table 1 refers to the number of pixels occupied by the grains in the grid by dividing the aligned grains into a grid. In the first embodiment, the whole image is 2.5 million pixels (about A4 size), and when 1000 grains are divided into regions, one grid is composed of 2500 pixels, and a standard grain is Is about 1500 pixels.
[0061]
Further, a function calculation data is obtained by performing a predetermined function calculation based on the raw data.
[0062]
There are types of function operation data as shown in Table 2 below.
[0063]
[Table 2]

[0064]
The controller 18 uses the above d. As the display target data to be displayed on the LCD monitor 18A. ~ F. Raw data, and g. ~ J. A part or all of the function calculation data is displayed in numerical data or a graphical display form (radar chart in this case).
[0065]
Based on the display contents, the quality of the inspector is determined for the determination items shown in Table 3 below.
[0066]
[Table 3]

[0067]
FIG. d, No. g-No. The function calculation data of j is displayed as a radar chart. This radar chart is displayed on a part of the LCD monitor 18A.
[0068]
As shown in FIG. 5, the image display form of the LCD monitor 18 </ b> A is, from the left, a grain display area 100 for displaying an image obtained by actually capturing a grain. A grid-like ruled line 102 that divides kernels into individual kernels is simultaneously displayed. The grain display area 100 can be scrolled independently, and a part of a total of 1000 grains (one measurement number) is displayed in the grain display area 100.
[0069]
On the other hand, on the right side of the grain display area 100, there is a feature data display area 104, the upper part of which is a graphic display area 106 for displaying the radar chart, and the lower part is a numerical data display area 108.
[0070]
The operation of the first embodiment will be described below.
[0071]
(Metering procedure)
With the tray 72 inserted, the tray 72 is put into the tray 52 of the hopper unit 44 by an operator. In this case, since the hopper unit 44 corresponds to the input port 48 of the cover 46 by the detection of the home position sensor 50, the operator only has to input an appropriate amount of grain to the input port 48.
[0072]
When the application kernel is input and an operation for prompting activation is performed, the light source 34B of the head unit 34 is turned on, the imaging element array 34A is activated, and the driving of the scanner motor 36 is started.
[0073]
The drive of the scanner motor 36 causes the head section 34 to start moving from a home position slightly upstream of the scan start position sensor 38.
[0074]
Therefore, immediately after the start of the movement, the position is detected by the scan start position sensor 38, the timer is started (t1), and when the head unit 34 is detected by the distance measuring position sensor 42, the timer is stopped (t2). . Thus, the movement time Δt (= t2−t1) of the head unit 34 from the scan start position sensor 38 to the distance measurement position sensor 42 can be obtained. Here, the moving speed v of the head unit 34 is determined based on the time Δt and the distance L from the scan start position sensor 38 to the distance measurement start position sensor 42 which is known in advance. _S Is calculated.
[0075]
The calculated moving speed v of the head unit 34 _S , The moving speed of the hopper unit 44 and the rotational linear speed of the rotary valve 54 are set, and the hopper unit 44 is moved in the sub-scanning direction while rotating the rotary valve 54.
[0076]
As a result, the grains as the inspection object are vertically and horizontally aligned on the tray 72 and are evenly scattered.
[0077]
In this state, photometry by the head unit 34 is performed. That is, the grains on the tray 72 are irradiated with light from the light source 34B, and the head unit 34 is moved while detecting the reflected light by the image sensor array 34A, so that the grains sprayed on a predetermined area of the tray 72. Capture individual data of the grain.
[0078]
When the photometry is completed, the tray 72 is ejected and moved, and the head unit 34 and the hopper unit 44 are returned to their initial positions (home positions) for standby for the next photometry.
[0079]
The controller 18 processes the received data, controls display on the LCD monitor 18A, and controls printing by the printer 20 when a print instruction is issued.
[0080]
(Display on LCD monitor)
Here, in the first embodiment, the display content on the LCD monitor 18A is determined by a predetermined function using raw data obtained by scanning of the head unit 34 and uniquely determined based on photometric data, and raw data. The feature amount data is composed of the calculated function operation data.
[0081]
The display items can be specified by the user, and the No. of raw data (No. a to No. f in Table 1) can be specified. d-No. f of the function operation data (No. g to No. k in Table 2). g-No. j can be selected.
[0082]
For example, as a radar chart, No. d, No. g-No. When the five items j are selected, a radar chart as shown in FIG. 5 is displayed. In addition, at this time, data not simultaneously displayed on the radar chart is displayed in numerical values in a table format.
[0083]
Further, in FIG. 4, data based on the national average is displayed on the radar chart in the graphic display area 106, and the comparison result with the national average is displayed in the numerical data / text message display area 108 using a comparative adjective. Displays a text message in the words that were used.
[0084]
For example, a text message such as "blue, slightly thinner and smaller than the average rice in the whole country" is displayed.
[0085]
The inspector (user) visually checks the display state on the LCD monitor 18A, and determines pass / fail of each of the judgment items (see Table 3).
[0086]
Since the feature data is raw data and function operation data uniquely determined from the photometric data, and the text message is only the comparison result (not the judgment) with the national average, at this point, the threshold value There is no difference that can occur when the pass / fail is automatically determined by the user.
[0087]
In other words, pass / fail judgment depends on the inspector (user), but in a state where the threshold value as the pass / fail judgment criterion is inappropriate, it is better than the automatic judgment to determine the production area and variety of the grain. If the inspector who is familiar with the difference visually judges the accurate data (feature amount data) and the text message for comparison with the national average, it is possible to determine the quality of the product with high accuracy .
[0088]
(Second embodiment)
Hereinafter, a second embodiment of the present invention will be described. In the second embodiment, the same components as those in the first embodiment are denoted by the same reference numerals, and the description of the configuration will be omitted.
[0089]
The feature of the second embodiment is that the threshold not set in the first embodiment can be set to an appropriate threshold based on the displayed feature data and the text message. It is in the point which did.
[0090]
In the first embodiment, the raw data or the function calculation data is displayed on the LCD monitor 18A as the grain characteristic amount data. In the second embodiment, the determination result of the determination item is displayed. It is characterized by doing.
[0091]
In order to judge the acceptability of this judgment item, a threshold value for the judgment parameter in each judgment item (see Table 3) is required.
[0092]
In the second embodiment, a default value is set and stored in advance for each determination item.
[0093]
Here, the user can select one of the default mode and the user setting mode when making a pass / fail judgment of the judgment item.
[0094]
When the default mode is selected, the stored default value is adopted as a threshold, and a pass / fail judgment is made.
[0095]
On the other hand, when the user setting mode is selected, for each determination item, a plurality of thresholds can be arbitrarily set for each production area and each kind, Pass / fail judgment of each judgment item is performed.
[0096]
In the user setting mode, it is possible to confirm the above-mentioned feature amount data in advance, and it is possible to set an appropriate threshold by looking at this feature amount data.
[0097]
In other words, the user sets the threshold value as needed, so that the threshold value suitable for a plurality of grain production areas is stored over time, and if necessary, the user can select from among them. This is a point that can be reused.
[0098]
In the present apparatus, in the user setting, since the setting can be performed with reference to the feature amount data and the text message, an appropriate threshold value in a relatively local area such as a production area gradually increases, and the apparatus is easy to use in the production area, and It is possible to set an appropriate threshold value.
[0099]
FIG. 6 shows a control block diagram for setting a threshold value and determining whether or not the threshold value is good in the controller 18 according to the second embodiment.
[0100]
The photometric data is sent to the feature data generator 110, where raw data (see Table 1) and function operation data (see Table 2) are generated. The photometric data is sent to the LCD control unit 114 via the image processing unit 112 as an image of each grain.
[0101]
The feature data generated by the feature data generator 110 is sent to the determination item selector 118 based on the instruction signal from the determination execution instructor 116.
[0102]
That is, the determination item selection unit 118 sequentially reads out the determination items (see Table 3) from the determination item memory 121, and reads out feature amount data necessary for the quality determination of the determination items from the feature amount data generation unit 110.
[0103]
The read feature amount data is sent to the determination unit 120, and a pass / fail determination is made by comparing the read feature amount data with a threshold value (default value or user set value) stored in the applied threshold value storage unit 122. . The pass / fail judgment result is sent to the LCD display control unit 114 and displayed on the LCD monitor 18A.
[0104]
Here, the threshold value to be applied as a comparison this time is selectively stored in the applied threshold value storage unit 122, and the threshold value is set by the threshold value setting unit 124. You.
[0105]
The threshold value setting unit 124 is provided with a default value memory 124A and a user setting threshold value memory 124B, and stores a standard threshold value corresponding to each setting item in advance. The default value is read out when a threshold value selecting unit 126 connected to the threshold value setting unit 124 performs a selection operation to use the default value, and is sent to the applied threshold value storage unit 122. Is done.
[0106]
On the other hand, when there is a selection operation for user setting in threshold selection section 126, the threshold stored in user setting threshold memory 124B (if a plurality of thresholds are stored, each ID (identification code) is assigned) and sent to the applied threshold storage unit 122.
[0107]
Here, a threshold can be additionally registered in the user-set threshold memory 124B by an instruction operation from the threshold creation instruction operation unit 127.
[0108]
In this case, a user setting signal is sent to the LCD display control unit 114 to instruct the display of the feature data and the text message.
[0109]
As a result, the feature amount data and the text message are displayed on the LCD monitor 18A (the display form may be the same as that of the first embodiment, or may be a numerical value in a table form), and the user can display the data. Based on the content, a threshold is input by operating the threshold input operation unit 128. A window screen 130 as shown in FIG. 7 is displayed on the LCD monitor 18A as a threshold input screen.
[0110]
When the threshold setting by the user is completed, the set threshold is stored in the user setting threshold memory 124B of the threshold setting unit 124 together with the ID.
[0111]
FIG. 8 shows each determination item, a procedure for setting a threshold value for the quality determination, and a quality determination procedure.
[0112]
The pass / fail judgment is performed in order from step 200, and finally the remaining particles are sized.
[0113]
Steps 200 and 202 are for judging the quality of the immature blue grains by using whiteness evaluation data (function operation data No. g) and blueness evaluation data (function operation data No. h). In step 200, it is determined whether or not the color is blue. If the determination is affirmative, the process proceeds to step 202 to determine whether or not the color is white. If a negative determination (not white) is made in this step 202, it is determined that the grain is immature blue (step 204), and if an affirmative determination (white) is made, it is determined that the rice is dead blue (step 206). If a negative determination is made in step 200, the process moves to the next step 208.
[0114]
Step 208 is a determination of the quality of the milky white grains, which is determined using the whiteness evaluation data (function operation data No. g). If an affirmative determination (white) is made in step 208, it is determined that it is a milky white grain (step 210). If a negative determination is made in step 208, the process proceeds to the next step 212.
[0115]
Step 212 is a pass / fail judgment of the white belly, which is determined using the white pixel ratio. The white pixel ratio is the raw data No. a, No. b, No. (R> threshold, G> threshold, and B> threshold number of pixels) / calculated value of the number of pixels per grain, and whether or not it is partially white Is determined. If an affirmative determination (partially white) is made in step 212, it is determined to be a belly white grain (step 214). If a negative determination is made in step 212, the process proceeds to step 216.
[0116]
Step 216 is a pass / fail judgment of the white dead grain (rice) using the whiteness evaluation data (function operation data No. g). If an affirmative determination (white) is made in step 216, it is determined that the product is white dead rice (step 218). If a negative determination is made in step 216, the process proceeds to step 220.
[0117]
Steps 220, 222, and 224 are for determining the quality of the crushed granules, and are performed using the size (raw data No. d) and the roundness data (function operation data No. j). If an affirmative determination (small) is made in step 220, it is determined that the granules are crushed (step 226). If a negative determination is made in step 220, the process proceeds to step 222.
[0118]
Also in the case where an affirmative determination (small) is made in step 222 and then an affirmative determination (round) is made in step 224, it is also determined that the granules are crushed (step 228).
[0119]
If a negative determination is made in steps 222 and 224, the process proceeds to step 230.
[0120]
Step 230 is the other damage determination, which is performed using the redness evaluation data (function operation data No. i). If an affirmative determination (red) is made in step 230, it is determined that the damage is other (step 232). If a negative determination is made in step 230, the process proceeds to step 234.
[0121]
In step 234, the quality of the colored particles is determined using color change pixel rate data (function operation data No. k). If an affirmative determination (partially red) is made in step 234, it is determined that the grain is a colored grain (step 236). If a negative determination is made in step 234, the process proceeds to step 238.
[0122]
Step 238 is for judging other immature grains by using the size (raw data No. d). If an affirmative determination (small) is made in step 238, the grain is determined to be another immature grain (step 240). If a negative determination is made in step 238, the process proceeds to step 242.
[0123]
Step 242 is for judging other immature grains using the blueness evaluation data (function operation data No. h). If an affirmative determination (blue) is made in step 242, it is determined that the grain is another immature grain (step 244). If a negative determination is made in step 242, the process proceeds to step 246.
[0124]
Step 246 is a determination of other immature grains, which is determined using redness evaluation data (function operation data No. i). If an affirmative determination (red) is made in step 246, the grain is determined to be another immature grain (step 248). If a negative determination is made in step 246, the process proceeds to step 250.
[0125]
Step 250 is a determination of other immature grains, which is determined using whiteness evaluation data (function operation data No. g). If an affirmative determination (white) is made in step 250, it is determined that the grain is another immature grain (step 252). If a negative determination is made in step 250, the process proceeds to step 254.
[0126]
Step 254 is a pass / fail judgment of the body split, which is determined using the color change pixel rate data (function operation data No. k). If an affirmative determination is made in step 254 (there is a partial color change), it is determined that the body is split (step 256). If a negative determination is made in step 254, the process proceeds to step 258, where it is determined that the grain is sized.
[0127]
In the pass / fail judgment results for each grain, a graph using the xy axes is used, for example, the size is plotted on the x-axis, the roundness is plotted on the y-axis, and the pass / fail judgment of other immature grains is graphically displayed. do it. The quality of each grain can be determined by pointing the pointer at a desired grain image (or in the area surrounded by the ruled line 102) in the grain display area 100 and clicking on it, and the determination result of the corresponding grain is graphically displayed. It is displayed in the display area 106. This is an advantageous means for grasping the properties of each grain.
[0128]
It is possible to know how the apparatus numerically grasps the properties of particles, such as colored particles and damaged particles, that may belong to items that affect the judgment even in a small amount.
[0129]
As described above, in the second embodiment, only one default value, which is a standard threshold value, is stored in each judgment item, and when this default value is inappropriate, the feature amount data and the text message While watching, the threshold is set by the user setting. For this reason, it is possible to set a threshold directly connected to a local area such as a production area, and there is no problem that an inappropriate threshold is set by mistake.
[0130]
In addition, when the default value is used, it is possible to judge the pass / fail with a certain degree of accuracy even if the production area and the variety are different, so that workability is good. On the other hand, in the case of user setting, since the threshold value is set while checking the user's skilled skills and feature amount data, it takes time to set the threshold value, but the threshold value with high accuracy Settings can be made.
[0131]
In the first and second embodiments, the kernels are automatically spread (aligned) on the tray using the handling unit 26. However, the handling unit 26 is removed and the kernels are removed. Even if the structure is such that grains are placed on the transparent glass plate 30 by manually placing the glass plate randomly using an alignment plate having through holes arranged vertically and horizontally for aligning the grains, Good.
[0132]
【The invention's effect】
As described above, according to the present invention, by directly displaying raw data and function calculation data measured by scanning as feature data, it is possible to perform absolute evaluation without being affected by differences in production areas, varieties, etc. Has an effect.
[Brief description of the drawings]
FIG. 1 is a perspective view showing an overview of an autoloader type discriminating apparatus according to a first embodiment.
FIG. 2 is a schematic diagram showing an internal configuration of the autoloader type discriminating apparatus according to the first embodiment.
FIG. 3 is a cross-sectional view illustrating a structure of the hopper unit according to the first embodiment.
FIG. 4 is a front view showing an example of a radar chart according to the first embodiment.
FIG. 5 is a front view showing a display screen of the LCD monitor according to the first embodiment.
FIG. 6 is a control block diagram for threshold setting and pass / fail determination according to a second embodiment.
FIG. 7 is a front view showing a threshold user setting window screen according to a second embodiment.
FIG. 8 is a flowchart showing each determination item, a procedure for setting a threshold value for quality determination, and a quality determination procedure.
[Explanation of symbols]
10 Autoloader type discriminator (granular inspection object state discriminator)
12, 14 box
16 case
18 Controller
18A LCD monitor (characteristic amount display means, judgment result display means)
20 Printer
22 Discriminator body
24 Scanner section
30 transparent glass plate (photometric surface)
32 line scanning type imaging unit
34 Head
36 Scanner motor
100 grain display area
102 Ruled Line
104 Feature data display area
106 Graphic display area (characteristic amount display means)
108 Numerical data / text message display area (text message display means)
110 Feature Data Generator
112 Image processing unit
114 LCD display control unit
116 judgment execution instructing unit
118 Judgment item selection section
120 Judgment unit
121 Judgment item memory
122 applied threshold storage
124 threshold setting section (threshold setting means)
124A Default value memory
124B User setting threshold memory
126 Threshold selector
127 Threshold value creation instruction operation unit
128 Threshold input operation unit
130 Window screen

Claims (5)

In a state where the granular test object is aligned on the flat bed photometric surface, the granular test object is scanned while irradiating light, and reflected light or transmitted light is measured to obtain photometric data. A granular inspection object state determination device that determines a finished state in aligned granular inspection object units,
A feature value for displaying raw data obtained by analyzing the photometric data of the granular test object, and function operation data obtained by calculating the raw data based on a predetermined function, as feature data of the granular test object; Display means;
Character message display means for displaying the result of comparing the average value of the feature amount data with a reference average value using an adjective text message,
Inspection device for inspecting the state of a granular object having: The feature quantity display means graphically displays a list of some or all of the feature quantity data composed of the raw data and the function calculation data together with the reference average value for each granular test object. 2. The apparatus for determining the state of a granular inspection object according to claim 1, wherein: Based on the feature amount data, blue immature grains, blue dead grains, milky white grains, belly white grains, white dead grains, immature grains, crushed grains, damaged grains, colored grains, determining means of determining the quality of the body grain,
3. The apparatus according to claim 1, further comprising a threshold value setting unit that can set a threshold value as a reference for the quality determination with reference to display states of the feature amount display unit and the text message display unit. The apparatus for determining the state of a granular inspection object according to the above description. 4. The apparatus according to claim 3, wherein the threshold value set by the threshold value setting means can be accumulated and stored together with an identification code. The raw data includes red (R) color separation data, green (G) color separation data, blue (B) color separation data, size data of each inspection object, and length data of each inspection object. , Width data of each inspection object,
The function calculation data includes whiteness evaluation data, blueness evaluation data, redness evaluation data, roundness data, and color change pixel rate data. The granular inspection object state determination device according to any one of claims 1 to 4.

Images