TWI721632B - Device and method for setting product detection threshold and storage medium - Google Patents
Device and method for setting product detection threshold and storage medium Download PDFInfo
- Publication number
- TWI721632B TWI721632B TW108140142A TW108140142A TWI721632B TW I721632 B TWI721632 B TW I721632B TW 108140142 A TW108140142 A TW 108140142A TW 108140142 A TW108140142 A TW 108140142A TW I721632 B TWI721632 B TW I721632B
- Authority
- TW
- Taiwan
- Prior art keywords
- threshold
- product
- initial
- judgment
- test
- Prior art date
Links
Images
Landscapes
- Testing Of Individual Semiconductor Devices (AREA)
Abstract
Description
本發明涉及檢測技術領域,尤其涉及一種應用於產品檢測之檢測閾值設定裝置、方法及電腦可讀取存儲介質。 The present invention relates to the field of detection technology, in particular to a detection threshold setting device, method and computer readable storage medium applied to product detection.
於新產品導入初期,待測物之閾值範圍通常較為嚴格,需藉由大量人工目檢結果及產線實際狀況,一步步將閾值修正到適當範圍。習知之運作方式由工程師根據產線實際狀況,多次來回檔適一段時間後才會確定最終之閾值,因此需投入較多之人力成本。比如,自動光學辨識(AOI)機台應用於SMT組裝線上,檢測電路板上之零件組裝後之品質狀況,或是檢查錫膏印刷後有否符合標準,產線工程師須設定每個待測物之閾值,若標準設定太嚴格,則假警報率過高;若標準設定太寬鬆,又會漏檢。 In the initial stage of the introduction of new products, the threshold range of the test object is usually strict, and it is necessary to correct the threshold step by step to an appropriate range through a large number of manual visual inspection results and the actual status of the production line. The conventional operation method is determined by the engineer according to the actual situation of the production line, and the final threshold will be determined after a certain period of time. Therefore, more labor costs are required. For example, the automatic optical identification (AOI) machine is applied to the SMT assembly line to check the quality of the parts on the circuit board after assembly, or to check whether the solder paste meets the standard after printing. The production line engineer must set each object to be tested If the standard setting is too strict, the false alarm rate will be too high; if the standard setting is too loose, the detection will be missed.
有鑑於此,有必要提供一種產品檢測閾值設定裝置、方法及電腦可讀取存儲介質,可給出合適之檢測建議閾值,使得檢測機台誤報率降到最低、效益達到最高。 In view of this, it is necessary to provide a product detection threshold setting device, method, and computer readable storage medium, which can give appropriate detection thresholds, so that the false alarm rate of the detection machine is minimized and the benefit is maximized.
本發明一實施方式提供一種產品檢測閾值設定方法,所述方法包括:獲取檢測設備對待測產品之產品參數進行檢測所設定之初始檢測閾值及所 述初始檢測閾值之設定方式;統計於所述初始檢測閾值下所述待測產品之第一判定數量、第二判定數量、第三判定數量及第四判定數量,其中所述第一判定數量為所述檢測設備判定為良品,人工複判亦為良品之數量,所述第二判定數量為所述檢測設備判定為良品,人工複判為不良品之數量,所述第三判定數量為所述檢測設備判定為不良品,人工複判為良品之數量,所述第四判定數量為所述檢測設備判定為不良品,人工複判亦為不良品之數量;若所述初始檢測閾值之設定方式為單邊閾值下界,則獲取被所述檢測設備判定為不良品之所有待測產品中之最小產品參數;將所述最小產品參數、所述初始檢測閾值及所述最小產品參數與所述初始檢測閾值之間之多個數值加入一集合;從所述集合中任意取出一元素設為試驗閾值,並統計於所述試驗閾值下所述待測產品之第一判定數量、第二判定數量、第三判定數量及第四判定數量;基於所述初始檢測閾值下所述待測產品之第一判定數量、第三判定數量及所述試驗閾值下所述待測產品之第一判定數量、第三判定數量計算所述試驗閾值之效益;重複從所述集合中任意取出一元素設為所述試驗閾值之步驟,直至所述集合為空集,以計算所述集合中每一元素之效益;及將所述集合中具有最大效益之元素作為所述檢測設備對所述待測產品進行檢測之建議閾值下界。 An embodiment of the present invention provides a method for setting a product detection threshold. The method includes: acquiring an initial detection threshold set by a detection device to detect product parameters of a product to be tested and all The setting method of the initial detection threshold; the first judgment quantity, the second judgment quantity, the third judgment quantity, and the fourth judgment quantity of the product to be tested under the initial detection threshold are counted, wherein the first judgment quantity is The detection equipment judged to be a good product, the manual re-judgment is also the quantity of good products, the second judged quantity is the quantity judged by the detection equipment to be good, and the manual re-judgment is the quantity of defective products, and the third judged quantity is the quantity The detection equipment determines the number of defective products and manual re-judgment is good. The fourth judgment quantity is the number of defective products judged by the testing equipment, and the manual re-judgment is also the number of defective products; if the initial detection threshold is set Is the lower bound of the unilateral threshold, then obtain the minimum product parameter among all the products to be tested that are judged as defective by the testing equipment; compare the minimum product parameter, the initial detection threshold, and the minimum product parameter with the initial The multiple values between the detection thresholds are added to a set; an element is randomly taken out of the set as the test threshold, and the first judgment quantity, the second judgment quantity, and the test product under the test threshold are counted. The third judgment quantity and the fourth judgment quantity; based on the first judgment quantity and the third judgment quantity of the product under test under the initial detection threshold, and the first judgment quantity and the first judgment quantity of the product under test under the test threshold 3. Calculate the benefit of the test threshold by determining the number; repeat the step of randomly removing an element from the set as the test threshold until the set is empty to calculate the benefit of each element in the set; And the element with the greatest benefit in the set is used as the lower bound of the recommended threshold for the detection device to detect the product under test.
優選地,所述方法還包括:若所述初始檢測閾值之設定方式為單邊閾值上界,則獲取被所述檢測設備判定為不良品之所有待測產品中之最大產品參數;將所述最大產品參數、所述初始檢測閾值及所述最大產品參數與所述初始檢測閾值之間之多個數值加入一集合;及將所述集合中具有最大效益之元素作為所述檢測設備對所述待測產品進行檢測之建議閾值上界。 Preferably, the method further includes: if the initial detection threshold is set as a unilateral threshold upper bound, obtaining the largest product parameter among all the products to be tested that are judged as defective by the detection equipment; The maximum product parameter, the initial detection threshold, and multiple values between the maximum product parameter and the initial detection threshold are added to a set; and the element with the greatest benefit in the set is used as the detection device to the The upper bound of the recommended threshold for testing the product to be tested.
優選地,所述將所述集合中具有最大效益之元素作為所述檢測設備對所述待測產品進行檢測之建議閾值上界之步驟還包括:若所述集合中具有最大效益之元素存於多個,則分別計算該多個元素與所述初始閾值之差值;從 該多個元素中選取與所述初始閾值差值最小之元素作為目標元素;及將所述目標元素作為所述檢測設備對所述待測產品進行檢測之建議閾值上界。 Preferably, the step of using the element with the greatest benefit in the set as the upper bound of the recommended threshold for the detection device to detect the product under test further includes: if the element with the greatest benefit in the set exists Multiple, calculate the difference between the multiple elements and the initial threshold; from The element with the smallest difference from the initial threshold is selected from the plurality of elements as the target element; and the target element is used as the recommended upper bound of the threshold for the detection device to detect the product under test.
優選地,所述將所述集合中具有最大效益之元素作為所述檢測設備對所述待測產品進行檢測之建議閾值下界之步驟包括:若所述集合中具有最大效益之元素存於多個,則分別計算該多個元素與所述初始閾值之差值;從該多個元素中選取與所述初始閾值差值最小之元素作為目標元素;及將所述目標元素作為所述檢測設備對所述待測產品進行檢測之建議閾值下界。 Preferably, the step of using the element with the greatest benefit in the set as the lower bound of the recommended threshold for the detection device to detect the product under test includes: if the element with the greatest benefit in the set exists in multiple , Calculate the difference between the plurality of elements and the initial threshold respectively; select the element with the smallest difference from the initial threshold from the plurality of elements as the target element; and use the target element as the detection device pair The lower bound of the recommended threshold for the product to be tested.
優選地,所述試驗閾值之效益藉由以下公式計算得到:BF=(TN’-TN)*COST1-(FN’-FN)*COST2,其中BF為所述試驗閾值之效益,TN’為於所述試驗閾值下所述待測產品之第一判定數量,TN為於所述初始檢測閾值下所述待測產品之第一判定數量,FN’為於所述試驗閾值下所述待測產品之第三判定數量,FN為於所述初始檢測閾值下所述待測產品之第三判定數量,COST1為所述檢測設備將良品判斷為不良品所帶來之成本,COST2為所述檢測設備將不良品判斷為良品所帶來之成本。 Preferably, the benefit of the test threshold is calculated by the following formula: BF=(TN'-TN)*COST1-(FN'-FN)*COST2, where BF is the benefit of the test threshold, and TN' is The first determined quantity of the product under test under the test threshold, TN is the first determined quantity of the product under test under the initial detection threshold, and FN' is the product under test under the test threshold The third judgment quantity, FN is the third judgment quantity of the product under test under the initial detection threshold, COST1 is the cost incurred by the detection equipment judging a good product as a defective product, and COST2 is the detection equipment The cost of judging defective products as good products.
本發明一實施方式提供一種產品檢測閾值設定方法,所述方法包括:獲取檢測設備對待測產品之產品參數進行檢測所設定之初始下界檢測閾值及初始上界檢測閾值;統計於所述初始下界檢測閾值下所述待測產品之第一判定數量、第二判定數量、第三判定數量及第四判定數量;獲取於所述初始下界檢測閾值下被所述檢測設備判定為不良品之所有待測產品中之最小產品參數;將所述最小產品參數、所述初始下界檢測閾值及所述最小產品參數與所述初始下界檢測閾值之間之多個數值加入第一集合;從所述第一集合中任意取出一元素設為第一試驗閾值,並統計於所述第一試驗閾值下所述待測產品之第一判定數量、第二判定數量、第三判定數量及第四判定數量;基於所述初始下界檢測閾值下所述待測產品之第一判定數量、第三判定數量及所述第一試驗閾值下所 述待測產品之第一判定數量、第三判定數量計算所述第一試驗閾值之效益;重複從所述第一集合中任意取出一元素設為所述第一試驗閾值之步驟,直至所述第一集合為空集,以計算所述第一集合中每一元素之效益;將所述第一集合中具有最大效益之元素作為所述檢測設備對所述待測產品進行檢測之建議閾值下界;統計於所述初始上界檢測閾值下所述待測產品之第一判定數量、第二判定數量、第三判定數量及第四判定數量;獲取於所述初始上界檢測閾值下被所述檢測設備判定為不良品之所有待測產品中之最大產品參數;將所述最大產品參數、所述初始上界檢測閾值及所述最大產品參數與所述初始上界檢測閾值之間之多個數值加入第二集合;從所述第二集合中任意取出一元素設為第二試驗閾值,並統計於所述第二試驗閾值下所述待測產品之第一判定數量、第二判定數量、第三判定數量及第四判定數量;基於所述初始上界檢測閾值下所述待測產品之第一判定數量、第三判定數量及所述第二試驗閾值下所述待測產品之第一判定數量、第三判定數量計算所述第二試驗閾值之效益;重複從所述第二集合中任意取出一元素設為所述第二試驗閾值之步驟,直至所述第二集合為空集,以計算所述第二集合中每一元素之效益;及將所述第二集合中具有最大效益之元素作為所述檢測設備對所述待測產品進行檢測之建議閾值上界;其中,所述第一判定數量為所述檢測設備判定為良品,人工複判亦為良品之數量,所述第二判定數量為所述檢測設備判定為良品,人工複判為不良品之數量,所述第三判定數量為所述檢測設備判定為不良品,人工複判為良品之數量,所述第四判定數量為所述檢測設備判定為不良品,人工複判亦為不良品之數量。 An embodiment of the present invention provides a method for setting a product detection threshold, the method comprising: acquiring an initial lower detection threshold and an initial upper detection threshold set by a detection device for detecting product parameters of a product to be tested; and collecting statistics on the initial lower detection The first judgment quantity, the second judgment quantity, the third judgment quantity, and the fourth judgment quantity of the product to be tested under the threshold; to obtain all the products to be tested that are judged as defective by the testing equipment under the initial lower detection threshold The smallest product parameter in the product; adding the smallest product parameter, the initial lower bound detection threshold, and multiple values between the smallest product parameter and the initial lower bound detection threshold to the first set; from the first set Take any element out of the set as the first test threshold, and count the first, second, third, and fourth judgments of the product under test under the first test threshold; The first judgment quantity, the third judgment quantity of the product to be tested under the initial lower detection threshold and the first test threshold The first judgment quantity and the third judgment quantity of the product to be tested calculate the benefit of the first test threshold; repeat the step of taking any element from the first set as the first test threshold until the The first set is an empty set to calculate the benefit of each element in the first set; the element with the greatest benefit in the first set is used as the lower bound of the recommended threshold for the detection equipment to detect the product under test ; Count the first, second, third, and fourth judgments of the product to be tested under the initial upper-bound detection threshold; obtain under the initial upper-bound detection threshold by the The largest product parameter among all the products to be tested that the testing equipment judges to be defective; the maximum product parameter, the initial upper detection threshold, and the multiple between the maximum product parameter and the initial upper detection threshold The value is added to the second set; an element is randomly selected from the second set as the second test threshold, and the first judgment quantity, the second judgment quantity, and the second judgment quantity of the product to be tested under the second test threshold are counted The third judgment quantity and the fourth judgment quantity; based on the first judgment quantity of the product to be tested under the initial upper detection threshold, the third judgment quantity and the first of the product under test under the second test threshold Determine the number of judgments and the third judgment number to calculate the benefit of the second test threshold; repeat the step of taking any element from the second set as the second test threshold until the second set is an empty set, To calculate the benefit of each element in the second set; and use the element with the greatest benefit in the second set as the upper bound of the recommended threshold for the detection equipment to detect the product under test; wherein, the The first judged quantity is the quantity judged by the testing equipment as good, and the manual re-judgment is also the quantity of good products. The second judged quantity is the quantity judged by the testing equipment as good and the manual re-judgment is bad. The judged quantity is the quantity judged by the inspection equipment as defective and the manual re-judgment is good, the fourth judged quantity is the quantity judged by the inspection equipment as defective, and the manual rejudgment is also the quantity of defective products.
優選地,所述將所述第一集合中具有最大效益之元素作為所述檢測設備對所述待測產品進行檢測之建議閾值下界之步驟包括:若所述第一集合中具有最大效益之元素存於多個,則分別計算該多個元素與所述初始下界檢測閾值之差值;及從該多個元素中選取與所述初始下界檢測閾值差值最小之元素 作為所述檢測設備對所述待測產品進行檢測之建議閾值下界。 Preferably, the step of using the element with the greatest benefit in the first set as the lower bound of the recommended threshold for the detection device to detect the product under test includes: if the element with the greatest benefit in the first set If stored in multiple, the difference between the multiple elements and the initial lower bound detection threshold is calculated separately; and the element with the smallest difference with the initial lower bound detection threshold is selected from the multiple elements As the lower bound of the recommended threshold for the detection device to detect the product under test.
優選地,所述將所述第二集合中具有最大效益之元素作為所述檢測設備對所述待測產品進行檢測之建議閾值上界之步驟包括:若所述第二集合中具有最大效益之元素存於多個,則分別計算該多個元素與所述初始上界檢測閾值之差值;從該多個元素中選取與所述初始上界檢測閾值差值最小之元素作為目標元素;及將所述目標元素作為所述檢測設備對所述待測產品進行檢測之建議閾值上界。 Preferably, the step of using the element with the greatest benefit in the second set as the upper bound of the recommended threshold for the detection device to detect the product under test includes: if the element in the second set has the greatest benefit If there are multiple elements, the difference between the multiple elements and the initial upper detection threshold is calculated separately; the element with the smallest difference from the initial upper detection threshold is selected from the multiple elements as the target element; and The target element is used as the upper bound of the recommended threshold for the detection device to detect the product to be tested.
本發明一實施方式提供一種產品檢測閾值設定裝置,所述裝置包括處理器及記憶體,所述記憶體上存儲有複數電腦程式,所述處理器用於執行記憶體中存儲之電腦程式時實現上述之產品檢測閾值設定方法之步驟。 An embodiment of the present invention provides a product detection threshold setting device. The device includes a processor and a memory. A plurality of computer programs are stored on the memory. The processor is used to execute the computer programs stored in the memory to achieve the above The steps of the product detection threshold setting method.
本發明一實施方式還提供一種電腦可讀取存儲介質,所述電腦可讀取存儲介質存儲有多條指令,多條所述指令可被一個或者多個處理器執行,以實現上述之產品檢測閾值設定方法之步驟。 An embodiment of the present invention also provides a computer-readable storage medium. The computer-readable storage medium stores a plurality of instructions, and the plurality of instructions can be executed by one or more processors to realize the above-mentioned product detection. Steps of threshold setting method.
與習知技術相比,上述產品檢測閾值設定裝置、方法及電腦可讀取存儲介質,藉由對設備判定為不良品之資料進行分析,可自動給出合適之檢測建議閾值,使得檢測機台誤報率降到最低、效益達到最高。 Compared with the prior art, the above-mentioned product detection threshold setting device, method, and computer readable storage medium can automatically give appropriate detection recommended thresholds by analyzing the equipment judged as defective products, so that the detection machine The false alarm rate is minimized and the benefit is maximized.
11:檢測設備 11: Testing equipment
13:待測產品 13: Product to be tested
10:記憶體 10: Memory
20:處理器 20: processor
30:產品檢測閾值設定程式 30: Product detection threshold setting program
101:第一獲取模組 101: The first acquisition module
102:統計模組 102: Statistics Module
103:第二獲取模組 103: The second acquisition module
104:加入模組 104: Add module
105:試驗模組 105: Test Module
106:計算模組 106: calculation module
107:建議模組 107: Suggested Module
100:產品檢測閾值設定裝置 100: Product detection threshold setting device
圖1是本發明一實施方式之待測產品之檢測環境示意圖。 Fig. 1 is a schematic diagram of a testing environment for a product to be tested according to an embodiment of the present invention.
圖2是本發明一實施方式之產品檢測閾值設定裝置之功能模組圖。 Fig. 2 is a functional module diagram of a product detection threshold setting device according to an embodiment of the present invention.
圖3是本發明一實施方式之產品檢測閾值設定程式之功能模組圖。 3 is a functional module diagram of a product detection threshold setting program according to an embodiment of the present invention.
圖4是本發明一實施方式之產品檢測閾值設定方法之流程圖。 Fig. 4 is a flowchart of a method for setting a product detection threshold according to an embodiment of the present invention.
請參閱圖1,檢測設備11用於對待測產品13進行檢測,以判斷待測產品13是良品還是不良品。所述檢測設備11可預先存儲有檢測標準,藉由檢測待測產品13之產品參數是否符合所述檢測標準來判斷待測產品13是良品還是不良品。比如,所述檢測設備11為AOI檢測機台,所述待測產品13為電路板,所述產品參數可是電路板之每一元器件之組裝狀況,或者錫膏印刷狀況。電路板中每一產品參數皆有一初始閾值,若量測值介於閾值內,則檢測設備11判斷其為良品;若量測值介於閾值外,則檢測設備11判斷其為不良品。經過所述檢測設備11檢測之電路板再經人工目檢後,複判結果共有以下四種情形:a).檢測設備11判定為良品,複判後亦為良品,其數量為TN;b).檢測設備11判定為良品,複判後為不良品,其數量為FN;c).檢測設備11判定為不良品,複判後為良品,其數量為FP;d).檢測設備11判定為不良品,複判後亦為不良品,其數量為TP。
Please refer to FIG. 1, the
請參閱圖2,為本發明產品檢測閾值設定裝置較佳實施例之示意圖。 Please refer to FIG. 2, which is a schematic diagram of a preferred embodiment of the product detection threshold setting device of the present invention.
所述產品檢測閾值設定裝置100包括記憶體10、處理器20以及存儲於所述記憶體10中並可於所述處理器20上運行之產品檢測閾值設定程式30。所述處理器20執行所述產品檢測閾值設定程式30時實現產品檢測閾值設定方法實施例中之步驟,例如圖4所示之步驟S400~S414。或者,所述處理器20執行所述產品檢測閾值設定程式30時實現產品檢測閾值設定程式實施例中各模組之功能,例如圖3中之模組101~107。
The product detection
於一實施方式中,所述產品檢測閾值設定裝置100可集成於所述檢測設備11中。
In one embodiment, the product detection
所述產品檢測閾值設定程式30可被分割成一個或多個模組,所述一個或者多個模組被存儲於所述記憶體10中,並由所述處理器20執行,以完成
本發明。所述一個或多個模組可是能夠完成特定功能之一系列電腦程式指令段,所述指令段用於描述所述產品檢測閾值設定程式30於所述產品檢測閾值設定裝置100中之執行過程。例如,所述產品檢測閾值設定程式30可被分割成圖3中之第一獲取模組101、統計模組102、第二獲取模組103、加入模組104、試驗模組105、計算模組106及建議模組107。各模組具體功能參見下圖3中各模組之功能。
The product detection
本領域技術人員可理解,所述示意圖僅是產品檢測閾值設定裝置100之示例,並不構成對產品檢測閾值設定裝置100之限定,可包括比圖示更多或更少之部件,或者組合某些部件,或者不同之部件,例如所述產品檢測閾值設定裝置100還可包括網路接入設備、匯流排等。
Those skilled in the art can understand that the schematic diagram is only an example of the product detection
所稱處理器20可是中央處理單元(Central Processing Unit,CPU),還可是其他通用處理器、數位訊號處理器(Digital Signal Processor,DSP)、專用積體電路(Application Specific Integrated Circuit,ASIC)、現成可程式設計閘陣列(Field-Programmable Gate Array,FPGA)或者其他可程式設計邏輯器件、分立門或者電晶體邏輯器件、分立硬體元件等。通用處理器可是微處理器或者所述處理器20亦可是任何常規之處理器等,所述處理器20可利用各種介面與匯流排連接產品檢測閾值設定裝置100之各個部分。
The so-called
所述記憶體10可用於存儲所述產品檢測閾值設定程式30與/或模組,所述處理器20藉由運行或執行存儲於所述記憶體10內之電腦程式與/或模組,以及調用存儲於記憶體10內之資料,實現所述產品檢測閾值設定裝置100之各種功能。所述記憶體10可包括高速隨機存取記憶體,還可包括非易失性記憶體,例如硬碟機、記憶體、插接式硬碟機,智慧存儲卡(Smart Media Card,SMC),安全數位(Secure Digital,SD)卡,快閃記憶體卡(Flash Card)、至少一個磁碟記憶體件、快閃記憶體器件、或其他非易失性固態記憶體件。
The
圖3為本發明產品檢測閾值設定程式較佳實施例之功能模組圖。 FIG. 3 is a functional module diagram of a preferred embodiment of the product detection threshold setting program of the present invention.
參閱圖3所示,產品檢測閾值設定程式30可包括第一獲取模組101、統計模組102、第二獲取模組103、加入模組104、試驗模組105、計算模組106及建議模組107。於一實施方式中,上述模組可為存儲於所述記憶體10中且可被所述處理器20調用執行之可程式化軟體指令。可理解之是,於其他實施方式中,上述模組亦可為固化於所述處理器20中之程式指令或固件(firmware)。
Referring to FIG. 3, the product detection
所述第一獲取模組101用於獲取所述檢測設備11對待測產品13之產品參數進行檢測所設定之初始檢測閾值及所述初始檢測閾值之設定方式。
The
於一實施方式中,所述初始檢測閾值可是測試人員根據以往測試經驗於所述檢測設備11中設定之檢測閾值。當初始檢測閾值被設定後,所述第一獲取模組101可獲取得到所述檢測設備11所設定之初始檢測閾值。所述初始檢測閾值之設定方式可包括三種:第一種為僅設置初始閾值下界LSL,該初始檢測閾值為[LSL,∞];第二種為僅設置初始閾值上界USL,該初始檢測閾值為[0,USL];第三種為設置了初始閾值下界LSL及初始閾值上界USL,該初始檢測閾值為[LSL,USL]。以下以僅設置初始閾值下界LSL為例進行說明。
In one embodiment, the initial detection threshold may be a detection threshold set in the
所述統計模組102用於統計於所述初始閾值下界下所述待測產品13之第一判定數量、第二判定數量、第三判定數量及第四判定數量。
The
於一實施方式中,所述第一判斷數量為TN,即所述檢測設備11判定為良品,人工複判亦為良品之數量,所述第二判定數量為FN,即所述檢測設備11判定為良品,人工複判為不良品之數量,所述第三判定數量為FP,即所述檢測設備11判定為不良品,人工複判為良品之數量,所述第四判定數量為TP,即所述檢測設備11判定為不良品,人工複判亦為不良品之數量。
In one embodiment, the first judgment quantity is TN, that is, the
於一實施方式中,由於每一所述待測產品13需要進行人工複判,所述第一判定數量、第二判定數量、第三判定數量及第四判定數量可由人工複
判統計得到,再錄入至所述檢測設備11,進而所述統計模組102可統計得到於所述初始閾值下界下所述待測產品13之第一判定數量、第二判定數量、第三判定數量及第四判定數量。
In one embodiment, since each of the products to be tested 13 needs to be manually re-judgmented, the first judgment quantity, the second judgment quantity, the third judgment quantity, and the fourth judgment quantity can be manually reviewed
The statistics are obtained, and then input into the
所述第二獲取模組103用於獲取被所述檢測設備11判定為不良品之所有待測產品13中之最小產品參數。
The
於一實施方式中,由於是以僅設置初始閾值下界LSL為例進行說明,則檢測設備11判為不良品之量測範圍為[0,LSL],判為良品之量測範圍為[LSL,∞]。於實際生產中,由於生產環境、生產參數等不能處於理想之生產狀態,所述檢測設備11可檢測得到多個設備判定之不良品,其產品參數之量測範圍於[0,LSL]之間,所述第二獲取模組103可獲取得到所有設備判定為不良品中之最小產品參數。舉例而言,LSL=10mm,設備判定為不良品之產品參數包括5mm、6mm、5mm、8mm、9mm、7mm,則待測產品13為不良品之最小產品參數為5mm。
In one embodiment, since only the initial threshold lower bound LSL is set as an example for description, the measuring range judged by the
所述加入模組104用於將所述最小產品參數、所述初始閾值下界及所述最小產品參數與所述初始閾值下界之間之多個數值加入一集合。
The adding
於一實施方式中,該多個數值可是所述最小產品參數與所述初始閾值下界之間之整數值,該多個數值還可是與所述初始閾值下界或所述最小產品參數構成一等差數列,等差數列之差值可根據實際需求進行設定。當所述加入模組104將所述最小產品參數、所述初始閾值下界及所述最小產品參數與所述初始閾值下界之間之多個數值加入所述集合後,所述集合即包括了多個元素。
In one embodiment, the multiple values may be integer values between the minimum product parameter and the lower bound of the initial threshold, and the multiple values may also constitute an arithmetic difference with the lower bound of the initial threshold or the minimum product parameter. The difference between the sequence and the arithmetic sequence can be set according to actual needs. When the adding
所述試驗模組105用於從所述集合中任意取出一元素設為試驗閾值,並統計於所述試驗閾值下所述待測產品13之第一判定數量、第二判定數量、第三判定數量及第四判定數量。
The
於一實施方式中,試驗模組105可從所述集合中任意取出一元素設為試驗閾值,並統計於所述試驗閾值下所述待測產品13之第一判定數量、第二判定數量、第三判定數量及第四判定數量。
In one embodiment, the
可理解所述試驗模組105可重複元素取出過程,直至所述集合為空集,即可統計得到所述集合中每一元素所對應之第一判定數量、第二判定數量、第三判定數量及第四判定數量。
It can be understood that the
所述計算模組106用於基於所述初始閾值下界下所述待測產品13之第一判定數量、第三判定數量及所述試驗閾值下所述待測產品13之第一判定數量、第三判定數量計算所述試驗閾值之效益。
The
於一實施方式中,當所述試驗模組105得到於所述試驗閾值下所述待測產品13之第一判定數量、第二判定數量、第三判定數量及第四判定數量時,所述計算模組106可基於所述初始閾值下界下所述待測產品13之第一判定數量、第三判定數量及所述試驗閾值下所述待測產品13之第一判定數量、第三判定數量計算所述試驗閾值之效益。可理解所述集合中之每一元素均是一試驗閾值,所述計算模組106可以相同之計算方式計算得到每一元素之效益。
In one embodiment, when the
於一實施方式中,所述試驗閾值之效益可根據實際需求設定計算公式,比如藉由以下公式計算得到:BF=(TN’-TN)*COST1-(FN’-FN)*COST2,其中BF為所述試驗閾值之效益,TN’為於所述試驗閾值下所述待測產品13之第一判定數量,TN為於所述初始閾值下界下所述待測產品13之第一判定數量,FN’為於所述試驗閾值下所述待測產品13之第三判定數量,FN為於所述初始閾值下界下所述待測產品13之第三判定數量,COST1為所述檢測設備11將良品判斷為不良品所帶來之成本,COST2為所述檢測設備11將不良品判斷為良品所帶來之成本。
In one embodiment, the benefit of the test threshold can be calculated according to actual needs. For example, it can be calculated by the following formula: BF=(TN'-TN)*COST1-(FN'-FN)*COST2, where BF For the benefit of the test threshold, TN' is the first determined quantity of the product under
所述建議模組107用於將所述集合中具有最大效益之元素作為所
述檢測設備11對所述待測產品13進行檢測之建議閾值下界。
The
於一實施方式中,當計算模組106計算得到所述集合中每一元素之效益時,所述建議模組107可從該些計算得到之效益中查找具有最大效益之元素,並將具有最大效益之元素作為所述檢測設備11對所述待測產品13進行檢測之建議閾值下界。
In one embodiment, when the
於一實施方式中,若產生最大效益之元素不僅一個,則所述建議模組107可取最接近所述初始閾值下界之元素作為所述建議閾值下界。具體地,若所述集合中具有最大效益之元素存於多個,則所述建議模組107可分別計算該多個元素與所述初始閾值下界之差值,再從該多個元素中選取與所述初始閾值下界差值最小之元素作為目標元素,最後將所述目標元素設定為所述檢測設備11對所述待測產品13進行檢測之建議閾值下界。
In one embodiment, if there is more than one element that produces the greatest benefit, the
於一實施方式中,若所述初始檢測閾值之設定方式為僅設置初始閾值上界USL。則檢測設備11判為不良品之量測範圍為[USL,∞],判為良品之量測範圍為[0,USL]。所述檢測設備11可檢測得到多個設備判定之不良品,其產品參數之量測範圍於[USL,∞]之間,所述第二獲取模組103可獲取得到所有設備判定為不良品中之最大產品參數。舉例而言,USL=10mm,設備判定為不良品之產品參數包括15mm、16mm、15mm、18mm、19mm、17mm,則待測產品13為不良品之最大產品參數為19mm。
In one embodiment, if the initial detection threshold is set in a manner that only the initial upper threshold USL is set. Then, the measurement range judged by the
於一實施方式中,所述加入模組104可將所述最大產品參數、所述初始閾值上界及所述最大產品參數與所述初始閾值上界之間之多個數值加入一集合。該多個數值可是所述最大產品參數與所述初始閾值上界之間之整數值,該多個數值還可是與所述初始閾值上界或所述最大產品參數構成一等差數列,等差數列之差值可根據實際需求進行設定。當所述加入模組104將所述最大產品參數、所述初始閾值上界及所述最大產品參數與所述初始閾值上界之間之多
個數值加入所述集合後,所述集合即包括了多個元素。所述試驗模組105可從所述集合中任意取出一元素設為試驗閾值,並統計於所述試驗閾值下所述待測產品13之第一判定數量、第二判定數量、第三判定數量及第四判定數量。
In one embodiment, the adding
於一實施方式中,試驗模組105同樣可從所述集合中任意取出一元素設為試驗閾值,並統計於所述試驗閾值下所述待測產品13之第一判定數量、第二判定數量、第三判定數量及第四判定數量。所述計算模組106可基於所述初始閾值上界下所述待測產品13之第一判定數量、第三判定數量及所述試驗閾值下所述待測產品13之第一判定數量、第三判定數量計算所述試驗閾值之效益。
In one embodiment, the
於一實施方式中,當所述試驗模組105得到於所述試驗閾值下所述待測產品13之第一判定數量、第二判定數量、第三判定數量及第四判定數量時,所述計算模組106可基於所述初始閾值上界下所述待測產品13之第一判定數量、第三判定數量及所述試驗閾值下所述待測產品13之第一判定數量、第三判定數量計算所述試驗閾值之效益。
In one embodiment, when the
於一實施方式中,當計算模組106計算得到所述集合中每一元素之效益時,所述建議模組107可從該些計算得到之效益中查找具有最大效益之元素,並將具有最大效益之元素作為所述檢測設備11對所述待測產品13進行檢測之建議閾值上界。
In one embodiment, when the
於一實施方式中,若產生最大效益之元素不僅一個,則所述建議模組107可取最接近所述初始閾值上界之元素作為所述建議閾值上界。具體地,若所述集合中具有最大效益之元素存於多個,則所述建議模組107可分別計算該多個元素與所述初始閾值上界之差值,再從該多個元素中選取與所述初始閾值上界差值最小之元素作為目標元素,最後將所述目標元素設定為所述檢測設備11對所述待測產品13進行檢測之建議閾值上界。
In one embodiment, if there is more than one element that produces the greatest benefit, the
於一實施方式中,若所述初始檢測閾值之設定方式為同時設定了初始閾值下界LSL及初始閾值上界USL,則檢測設備11判為不良品之量測範圍為[USL,∞]及[0,USL],判為良品之量測範圍為[LSL,USL]。基於上述確定建議閾值下界之方式,對在於[0,USL]區間設備判定之不良品,可找到所述檢測設備11對所述待測產品13進行檢測之建議閾值下界,於此不再重複敘述。基於上述確定建議閾值上界之方式,對在於[USL,∞]區間設備判定之不良品,可找到所述檢測設備11對所述待測產品13進行檢測之建議閾值上界,於此不再重複敘述。
In one embodiment, if the initial detection threshold is set in a way that the initial lower threshold LSL and the initial upper threshold USL are set at the same time, the measurement range of the
圖4為本發明一實施方式中產品檢測閾值設定方法之流程圖。根據不同之需求,所述流程圖中步驟之順序可改變,某些步驟可省略。 4 is a flowchart of a method for setting a product detection threshold in an embodiment of the present invention. According to different needs, the order of the steps in the flowchart can be changed, and some steps can be omitted.
步驟S400,獲取檢測設備11對待測產品13之產品參數進行檢測所設定之初始檢測閾值及所述初始檢測閾值之設定方式。
In step S400, the initial detection threshold set by the
步驟S402,統計於所述初始檢測閾值下所述待測產品13之第一判定數量、第二判定數量、第三判定數量及第四判定數量。
Step S402: Count the first judgment quantity, the second judgment quantity, the third judgment quantity, and the fourth judgment quantity of the product under
步驟S404,若所述初始檢測閾值之設定方式為單邊閾值下界,則獲取被所述檢測設備11判定為不良品之所有待測產品13中之最小產品參數。
In step S404, if the setting method of the initial detection threshold is the lower bound of the unilateral threshold, obtain the minimum product parameter of all the products under
步驟S406,將所述最小產品參數、所述初始檢測閾值及所述最小產品參數與所述初始檢測閾值之間之多個數值加入一集合。 In step S406, the minimum product parameter, the initial detection threshold, and multiple values between the minimum product parameter and the initial detection threshold are added to a set.
步驟S408,從所述集合中任意取出一元素設為試驗閾值,並統計於所述試驗閾值下所述待測產品13之第一判定數量、第二判定數量、第三判定數量及第四判定數量。
Step S408: Take any element from the set as a test threshold, and count the first, second, third, and fourth judgments of the product under
步驟S410,基於所述初始檢測閾值下所述待測產品13之第一判定數量、第三判定數量及所述試驗閾值下所述待測產品13之第一判定數量、第三判定數量計算所述試驗閾值之效益。
Step S410, based on the first judgment quantity and the third judgment quantity of the product under
步驟S412,重複從所述集合中任意取出一元素設為所述試驗閾值之步驟,直至所述集合為空集,以計算所述集合中每一元素之效益。 Step S412: Repeat the step of randomly extracting an element from the set as the test threshold until the set is empty, so as to calculate the benefit of each element in the set.
步驟S414,將所述集合中具有最大效益之元素作為所述檢測設備11對所述待測產品13進行檢測之建議閾值下界。
In step S414, the element with the greatest benefit in the set is used as the lower bound of the recommended threshold for the
上述產品檢測閾值設定裝置、方法及電腦可讀取存儲介質,藉由對設備判定為不良品之資料進行分析,可自動給出合適之檢測建議閾值,使得檢測機台誤報率降到最低、效益達到最高。 The above-mentioned product detection threshold setting device, method, and computer readable storage medium can automatically give appropriate detection thresholds by analyzing the equipment judged to be defective products, so that the false alarm rate of the detection machine is minimized and beneficial Reach the highest.
綜上所述,本發明符合發明專利要件,爰依法提出專利申請。惟,以上所述者僅為本發明之較佳實施方式,本發明之範圍並不以上述實施方式為限,舉凡熟悉本案技藝之人士爰依本發明之精神所作之等效修飾或變化,皆應涵蓋於以下申請專利範圍內。 In summary, the present invention meets the requirements of an invention patent, and Yan filed a patent application in accordance with the law. However, the above are only the preferred embodiments of the present invention, and the scope of the present invention is not limited to the above embodiments. Anyone familiar with the art of the present case makes equivalent modifications or changes based on the spirit of the present invention. Should be covered in the scope of the following patent applications.
Claims (10)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW108140142A TWI721632B (en) | 2019-11-05 | 2019-11-05 | Device and method for setting product detection threshold and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW108140142A TWI721632B (en) | 2019-11-05 | 2019-11-05 | Device and method for setting product detection threshold and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
TWI721632B true TWI721632B (en) | 2021-03-11 |
TW202119144A TW202119144A (en) | 2021-05-16 |
Family
ID=76035931
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW108140142A TWI721632B (en) | 2019-11-05 | 2019-11-05 | Device and method for setting product detection threshold and storage medium |
Country Status (1)
Country | Link |
---|---|
TW (1) | TWI721632B (en) |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102138068A (en) * | 2008-09-09 | 2011-07-27 | 玛机统丽公司 | Visual examination apparatus |
CN103226470A (en) * | 2013-03-21 | 2013-07-31 | 北京神州绿盟信息安全科技股份有限公司 | Method and device for determining weighted value of check item based on BVS (benchmark verification system) |
WO2016035147A1 (en) * | 2014-09-02 | 2016-03-10 | 株式会社ニコン | Measurement processing device, measurement processing method, measurement processing program, and structure production method |
CN205879493U (en) * | 2016-06-12 | 2017-01-11 | 常州星宇车灯股份有限公司 | On -line measuring system of car light lens |
CN207423830U (en) * | 2017-10-27 | 2018-05-29 | 深圳精创视觉科技有限公司 | The automatic indication device of optical film shortcoming |
WO2019159426A1 (en) * | 2018-02-14 | 2019-08-22 | 株式会社Screenホールディングス | Inspection device, inspection method, tablet printing device, and tablet printing method |
WO2019176990A1 (en) * | 2018-03-13 | 2019-09-19 | オムロン株式会社 | Inspection device, image discriminating device, discriminating device, inspection method, and inspection program |
-
2019
- 2019-11-05 TW TW108140142A patent/TWI721632B/en active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102138068A (en) * | 2008-09-09 | 2011-07-27 | 玛机统丽公司 | Visual examination apparatus |
CN103226470A (en) * | 2013-03-21 | 2013-07-31 | 北京神州绿盟信息安全科技股份有限公司 | Method and device for determining weighted value of check item based on BVS (benchmark verification system) |
WO2016035147A1 (en) * | 2014-09-02 | 2016-03-10 | 株式会社ニコン | Measurement processing device, measurement processing method, measurement processing program, and structure production method |
CN205879493U (en) * | 2016-06-12 | 2017-01-11 | 常州星宇车灯股份有限公司 | On -line measuring system of car light lens |
CN207423830U (en) * | 2017-10-27 | 2018-05-29 | 深圳精创视觉科技有限公司 | The automatic indication device of optical film shortcoming |
WO2019159426A1 (en) * | 2018-02-14 | 2019-08-22 | 株式会社Screenホールディングス | Inspection device, inspection method, tablet printing device, and tablet printing method |
WO2019176990A1 (en) * | 2018-03-13 | 2019-09-19 | オムロン株式会社 | Inspection device, image discriminating device, discriminating device, inspection method, and inspection program |
Also Published As
Publication number | Publication date |
---|---|
TW202119144A (en) | 2021-05-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111949941B (en) | Equipment fault detection method, system, device and storage medium | |
US10104428B2 (en) | Video playing detection method and apparatus | |
CN107263830B (en) | Forming system | |
US7557598B2 (en) | Method of inspecting quiescent power supply current in semiconductor integrated circuit and device for executing the method | |
TWI721632B (en) | Device and method for setting product detection threshold and storage medium | |
CN112785100B (en) | Product detection threshold setting device, method and computer readable storage medium | |
CN103632232A (en) | Method and equipment for detecting products | |
CN107896232B (en) | IP address evaluation method and device | |
CN108073518A (en) | A kind of testing case management and device | |
CN113870754B (en) | Method and system for judging defects of panel detection electronic signals | |
CN107656187B (en) | Differential line test information determining method and device | |
CN114387249A (en) | Detection method, detection device, equipment and storage medium | |
JP5155938B2 (en) | Pattern contour detection method | |
CN112801523A (en) | Cigarette quality detection method, device and system and storage medium | |
JP5906965B2 (en) | In-vehicle ECU evaluation system | |
CN113295546A (en) | FPC microcircuit bending damage degree verification method and device and electronic equipment | |
CN110969402A (en) | Standard man-hour measuring device, method and electronic equipment | |
CN112907684B (en) | Humidity detection method, device, equipment and medium | |
TWI754304B (en) | Defect analyzing method and device, electronic device, and computer-readable storage medium | |
KR20150116086A (en) | Image acquisition method for avoiding vibration of Vision examination equipment | |
CN115080432A (en) | Method and device for determining software test progress and electronic equipment | |
CN117723550B (en) | Detection method and device of product quality detection equipment and electronic equipment | |
CN114676387A (en) | Data statistical analysis method, system, readable storage medium and computer equipment | |
CN115435749A (en) | Method and device for measuring inclination angle of bonding finger, terminal and storage medium | |
JP2005301617A (en) | Abnormality detection method |