TWM606206U - Intelligent device for optimizing parameters automatically - Google Patents
Intelligent device for optimizing parameters automatically Download PDFInfo
- Publication number
- TWM606206U TWM606206U TW109212467U TW109212467U TWM606206U TW M606206 U TWM606206 U TW M606206U TW 109212467 U TW109212467 U TW 109212467U TW 109212467 U TW109212467 U TW 109212467U TW M606206 U TWM606206 U TW M606206U
- Authority
- TW
- Taiwan
- Prior art keywords
- value
- parameter
- true
- detection type
- algorithm detection
- Prior art date
Links
Images
Landscapes
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
一種自動優化參數的智能裝置包括光學裝置、視覺檢測裝置、人工智慧裝置與參數調整裝置。人工智慧裝置設置有參數設定器、真假瑕疵數量判斷器與處理器,處理器連接至參數設定器與真假瑕疵數量判斷器。參數調整裝置設置有自動調整器。視覺檢測裝置將一待測產品之一第N個檢驗位置輸入至視覺檢測裝置。人工智慧裝置透過參數設定器根據待測產品之第N個檢驗位置設定演算法檢測種類之值作為第一參數值並且透過真假瑕疵數量判斷器來判斷第一參數值之真瑕疵數量與假瑕疵數量。An intelligent device for automatically optimizing parameters includes an optical device, a visual detection device, an artificial intelligence device and a parameter adjustment device. The artificial intelligence device is provided with a parameter setter, a true and false defect number determiner, and a processor, and the processor is connected to the parameter setter and the true and false defect number determiner. The parameter adjustment device is equipped with an automatic adjuster. The visual inspection device inputs the Nth inspection position of a product to be tested to the visual inspection device. The artificial intelligence device sets the value of the algorithm detection type as the first parameter value through the parameter setter according to the Nth inspection position of the product to be tested, and judges the true and false defects of the first parameter value through the true and false defect number determiner Quantity.
Description
一種視覺檢測裝置之參數設定,尤指一種關於自動優化參數的智能裝置。A parameter setting of a visual inspection device, especially an intelligent device for automatically optimizing parameters.
電子產品中的電路板上各種元件之品質及外觀檢測,為電路板於製造與檢驗過程中的重要步驟,為了使檢測更加準確,目前大多使用視覺檢測裝置,以影像方式進行檢測,現有的檢測方式係將電路板的標準檢測數值(電路板佈局)輸入至視覺檢測裝置,讓視覺檢測裝置的操作者,依據其所累積之經驗來調整該標準檢測數值,但此種做法的可靠度相當的低,常會發生合格的產品被判定為不良品,此時操作者就必須再次調整標準檢測數值,讓其他合格的產品也能通過檢測,此種方式即便是相當有經驗的操作者,也需要人工花時間去反覆測試,才能調整出適合的標準檢測數值,讓假缺點降低,且可以抓出真瑕疵點。The quality and appearance inspection of various components on the circuit board in electronic products is an important step in the manufacturing and inspection process of the circuit board. In order to make the inspection more accurate, most of the visual inspection devices are currently used to inspect by imaging. The existing inspection The method is to input the standard inspection value of the circuit board (circuit board layout) into the visual inspection device, allowing the operator of the visual inspection device to adjust the standard inspection value based on their accumulated experience, but this approach is quite reliable Low, it often happens that qualified products are judged as defective products. At this time, the operator must adjust the standard test value again so that other qualified products can pass the test. This method requires labor even for quite experienced operators. It takes time to repeat the test to adjust the appropriate standard detection value, reduce false defects, and catch true defects.
本創作提出一種自動優化參數的智能裝置,自動優化參數的智能裝置包括光學裝置、視覺檢測裝置、人工智慧裝置與參數調整裝置。視覺檢測裝置連接至光學裝置,人工智慧裝置連接至視覺檢測裝置並且參數調整裝置連接至人工智慧裝置。人工智慧裝置設置有參數設定器、真假瑕疵數量判斷器與處理器,處理器連接至參數設定器與真假瑕疵數量判斷器。參數調整裝置設置有自動調整器,自動調整器連接至處理器。視覺檢測裝置將待測產品之第N個檢驗位置輸入至視覺檢測裝置。人工智慧裝置透過參數設定器根據待測產品之第N個檢驗位置設定演算法檢測種類之值作為第一參數值並且透過真假瑕疵數量判斷器來判斷第一參數值之真瑕疵數量與假瑕疵數量。參數調整裝置透過自動調整器來自動調整演算法檢測種類之值為下一個參數值,其中人工智慧裝置透過處理器找到對應於真瑕疵數量最多且假瑕疵數量最少之演算法檢測種類之值作為最佳參數值,其中N大於一之正整數。This creation proposes an intelligent device that automatically optimizes parameters. The intelligent device that automatically optimizes parameters includes an optical device, a visual inspection device, an artificial intelligence device, and a parameter adjustment device. The visual inspection device is connected to the optical device, the artificial intelligence device is connected to the visual inspection device and the parameter adjustment device is connected to the artificial intelligence device. The artificial intelligence device is provided with a parameter setter, a true and false defect number determiner, and a processor, and the processor is connected to the parameter setter and the true and false defect number determiner. The parameter adjustment device is provided with an automatic adjuster, and the automatic adjuster is connected to the processor. The visual inspection device inputs the Nth inspection position of the product to be tested to the visual inspection device. The artificial intelligence device sets the value of the algorithm detection type as the first parameter value through the parameter setter according to the Nth inspection position of the product to be tested, and judges the true and false defects of the first parameter value through the true and false defect number determiner Quantity. The parameter adjustment device automatically adjusts the value of the algorithm detection type through the automatic adjuster to the next parameter value. The artificial intelligence device uses the processor to find the value corresponding to the algorithm detection type with the largest number of true defects and the smallest number of false defects as the most Best parameter value, where N is a positive integer greater than one.
在本創作之一實施例中,演算法檢測種類為外型輪廓、位置座標、顏色或亮度,並且演算法檢測種類之值為外型輪廓數值、位置座標數值、顏色數值或亮度數值。In an embodiment of this creation, the algorithm detection type is contour, position coordinate, color or brightness, and the value of the algorithm detection type is the contour value, position coordinate value, color value, or brightness value.
在本創作之一實施例中,自動調整器在自動調整演算法檢測種類之值為下一個參數值之步驟中,係在下限值往上限值之範圍中進行調整。In an embodiment of this creation, the automatic adjuster adjusts the value of the type detected by the automatic adjustment algorithm in the range of the lower limit to the upper limit in the step of the next parameter value.
在本創作之一實施例中,處理器在判斷演算法檢測種類之值是否調整完畢之步驟中,為判斷演算法檢測種類之值是否已在下限值與上限值之範圍內都進行過數值調整。In an embodiment of this creation, in the step of judging whether the value of the algorithm detection type is adjusted by the processor, the value of the algorithm detection type is determined to be within the range of the lower limit and the upper limit. Adjustment.
在本創作之一實施例中,每一個演算法檢測種類之值對應於一組關於真瑕疵數量與假瑕疵數量之數據。In an embodiment of this creation, the value of each algorithm detection category corresponds to a set of data about the number of true defects and the number of false defects.
為達上述目的,本創作之自動優化參數的智能裝置,能夠調整出適合的標準檢測數值,讓假瑕疵點降低,且可以抓出真瑕疵點。In order to achieve the above-mentioned purpose, the intelligent device that automatically optimizes the parameters of this creation can adjust the appropriate standard detection value, so that the false defects can be reduced, and the real defects can be caught.
底下藉由具體實施例詳加說明,當更容易瞭解本創作之目的、技術內容、特點及其所達成之功效。The following detailed descriptions are given by specific examples, and it will be easier to understand the purpose, technical content, characteristics and effects of this creation.
為能解決現有電路板上各種元件之品質及外觀檢測的問題,創作人經過多年的研究及開發,據以改善現有產品的缺點,後續將詳細介紹本創作如何以一種視自動優化參數的智能裝置來達到最有效率的功能訴求。In order to solve the problems of quality and appearance inspection of various components on existing circuit boards, the creators have gone through years of research and development to improve the shortcomings of existing products. The follow-up will introduce in detail how this creation uses a smart device that automatically optimizes parameters. To achieve the most efficient functional requirements.
請參閱第一圖,第一圖係為本創作的視覺檢測裝置之參數設定方法之流程圖。如圖一所示,視覺檢測裝置之參數設定方法100包括以下步驟:將待測產品之第N個檢驗位置輸入至視覺檢測裝置(步驟S110);根據待測產品之第N個檢驗位置設定演算法檢測種類之值作為第一參數值(步驟S120);判斷第一參數值之真瑕疵數量與假瑕疵數量(步驟S130);自動調整演算法檢測種類之值為下一個參數值(步驟S140);計算D步驟之真瑕疵數量與假瑕疵數量(步驟S150);判斷演算法檢測種類之值是否調整完畢(步驟S160);找到對應於真瑕疵數量最多且假瑕疵數量最少之演算法檢測種類之值作為最佳參數值(步驟S170),其中N大於一之正整數。Please refer to the first figure. The first figure is a flowchart of the parameter setting method of the visual inspection device created. As shown in Figure 1, the
須要注意的是,上述演算法檢測種類為外型輪廓、位置座標、顏色或亮度,但並不以這四種為限。另外,演算法檢測種類之值為外型輪廓數值、位置座標數值、顏色數值或亮度數值。此外,在自動調整演算法檢測種類之值為下一個參數值之步驟中(亦即步驟S140),係在下限值往上限值之範圍中進行調整(逐步上調)或在上限值往下限值之範圍中進行調整(逐步下調)。在判斷演算法檢測種類之值是否調整完畢之步驟中(亦即步驟S160),為判斷演算法檢測種類之值是否已在下限值與上限值之範圍內都進行過數值調整,其中數值調整的間距可為設計者所設定。每一演算法檢測種類之值對應於一組關於真瑕疵數量與假瑕疵數量之數據。上述所謂之上限值與下限值為設計者所設定。It should be noted that the detection types of the above algorithms are contour, position coordinates, color or brightness, but they are not limited to these four. In addition, the value of the algorithm detection type is the appearance contour value, the position coordinate value, the color value or the brightness value. In addition, in the step where the value of the detection type of the automatic adjustment algorithm is the next parameter value (i.e., step S140), the adjustment is performed within the range of the lower limit value to the upper limit value (gradual increase) or the upper limit value is lowered Adjust within the range of the limit (gradual downward adjustment). In the step of judging whether the value of the algorithm detection type has been adjusted (that is, step S160), to determine whether the value of the algorithm detection type has been adjusted within the range of the lower limit and the upper limit, the value adjustment is performed The spacing can be set by the designer. The value of each algorithm detection category corresponds to a set of data about the number of true defects and the number of false defects. The above-mentioned upper limit and lower limit are set by the designer.
詳細來說,請同時參照第一圖至第三圖,第二圖係為本創作的自動優化參數的智能裝置之區塊示意圖。第三圖係為本創作的檢測待測產品之示意圖。運用視覺檢測裝置之參數設定方法之自動優化參數的智能裝置200包括視覺檢測裝置220、光學裝置230、人工智慧裝置240與參數調整裝置250。人工智慧裝置240設置有參數設定器242、真假瑕疵數量判斷器244與處理器246。參數調整裝置設置250有自動調整器252。視覺檢測裝置220連接至光學裝置230,人工智慧裝置240連接至視覺檢測裝置220。參數調整裝置250連接至人工智慧裝置240。處理器246連接至參數設定器242與真假瑕疵數量判斷器244。自動調整器252連接至處理器246。光學裝置220用以對待測產品210進行拍照。人工智慧裝置240透過參數設定器242根據待測產品210之第N個檢驗位置設定演算法檢測種類之值作為第一參數值並且透過真假瑕疵數量判斷器244來判斷第一參數值之真瑕疵數量與假瑕疵數量。參數調整裝置250透過自動調整器252來自動調整演算法檢測種類之值為下一個參數值,其中人工智慧裝置240透過處理器246來找到對應於真瑕疵數量最多且假瑕疵數量最少之演算法檢測種類之值作為最佳參數值,其中N為大於一之正整數。本創作所提出之解決方案為針對現有電路板上各種元件之品質及外觀檢測的問題,其工作機制詳述如下。In detail, please refer to the first to third figures at the same time. The second figure is a block diagram of the intelligent device that automatically optimizes parameters. The third picture is a schematic diagram of the product under test created by this author. The
光學裝置230會對一個待測產品210進行拍照,之後會針對待測產品210之多個檢驗位置(W、X、Y與Z)進行檢驗,在此以四個檢驗位置(相當於N為1至4)為例,但不以此為限。首先,先針對第一個檢驗位置W進行檢驗,將待測產品210之第一個檢驗位置(在此為檢驗位置W)輸入至視覺檢測裝置220。之後,視覺檢測裝置220會根據待測產品210之第一個檢驗位置W設定演算法檢測種類之值作為一第一參數值,其中第一檢驗位置W所檢驗之演算法檢測種類之值為外型輪廓數值、位置座標數值、顏色數值或亮度數值,在此假設先檢驗之演算法檢測種類為亮度並且設定亮度之演算法檢測種類之值以作為第一參數值。接下來,透過人工智慧裝置240,來檢視第一個檢驗位置W中進而判斷第一參數值之真瑕疵數量與假瑕疵數量。舉例來說,在第一個檢驗位置W檢驗亮度時,所得到之第一個真瑕疵數量為8個且第一個假瑕疵數量為15個。之後,利用回饋修正之機制,透過參數調整裝置250來自動調整改變演算法檢測種類之值以獲得下一個亮度參數值(在此,以亮度值為例),再利用視覺檢測裝置220來計算第一個檢驗位置W中另一個亮度值之真瑕疵數量與假瑕疵數量,例如此時的真瑕疵數量為12個且假瑕疵數量為3個。接下來,透過人工智慧裝置240來進行覆判且計算,判斷第一個檢驗位置W之演算法檢測種類之值(在此調整之參數值為亮度值)是否已經調整完畢,亦即為判斷演算法檢測種類之值是否已在下限值與上限值之範圍內都進行過數值調整。The
如果人工智慧裝置240判斷尚未調整結束,則繼續透過參數調整裝置250來自動調整改變演算法檢測種類之值且進行後續步驟。如果人工智慧裝置240判斷已經調整結束,則透過人工智慧裝置240,找到對應於真瑕疵數量最多且假瑕疵數量最少之演算法檢測種類之值作為一最佳參數值。在上面的工作機制中,每一演算法檢測種類之值對應於一組關於真瑕疵數量與假瑕疵數量之數據。例如,本實施例在第一檢驗位置所要調整之參數值為亮度值,且總共調整了15次,之後人工智慧裝置240總結計算後發現第二次的數據是相對優化的數據,所以以第二次之數據做為亮度之最佳參數值。If the
在結束亮度之優化後,在同樣的第一檢驗位置W,自動優化參數的智能裝置200會開始來調整演算法檢測種類,例如外型輪廓、位置座標或顏色。透過以上機制來檢測、設定初始值、回饋且調整演算法檢測種類之值、總結並找出相對優化之參數值,會得到相對優化之一組數據。也就是說,分別依據上述之機制來進行調整演算法檢測種類之值,以分別獲得外型輪廓之最佳值、位置座標之最佳值或顏色之最佳值。藉此,以完成待測產品210之第一檢驗位置W之相關優化作業。After the brightness optimization is finished, at the same first inspection position W, the
接下來,會針對第二個檢驗位置X進行檢驗。將待測產品210之第二個檢驗位置(在此為檢驗位置X)輸入至視覺檢測裝置220。後續的工作機制如上所述,其餘檢驗位置Y與X也同理。因此,本創作就是依序在每一個檢驗位置上,進行演算法檢測種類之值之最佳化,此最佳化之標準為使得對應於真瑕疵數量最多且假瑕疵數量最少之演算法檢測種類之值為最佳參數值。Next, inspection will be carried out for the second inspection position X. The second inspection position (here, inspection position X) of the
綜上,本創作之視覺檢測裝置之自動優化參數的智能裝置,能夠調整出適合的標準檢測數值,讓假缺點降低,且可以抓出真瑕疵點。In summary, the intelligent device that automatically optimizes the parameters of the visual inspection device created by this creation can adjust the appropriate standard inspection value, reduce false defects, and catch true defects.
唯以上所述者,僅為本創作之較佳實施例而已,並非用來限定本創作實施之範圍。故即凡依本創作申請範圍所述之特徵及精神所為之均等變化或修飾,均應包括於本創作之申請專利範圍內。Only the above are only preferred embodiments of this creation, and are not used to limit the scope of implementation of this creation. Therefore, all equivalent changes or modifications made in accordance with the characteristics and spirit of the application scope of this creation shall be included in the scope of patent application of this creation.
100:視覺檢測裝置之參數設定方法 S110、S120、S130、S140、S150、S160、S170:步驟 200:自動優化參數的智能裝置 210:待測產品 220:視覺檢測裝置 230:光學裝置 240:人工智慧裝置 242:參數設定器 244:真假瑕疵數量判斷器 246:處理器 250:參數調整裝置 252:自動調整器 W、X、Y、Z:檢驗位置 100: Parameter setting method of visual inspection device S110, S120, S130, S140, S150, S160, S170: steps 200: Smart device that automatically optimizes parameters 210: product to be tested 220: Visual inspection device 230: optical device 240: Artificial Intelligence Device 242: Parameter Setter 244: True and False Defects Number Judge 246: processor 250: Parameter adjustment device 252: automatic adjuster W, X, Y, Z: inspection position
第一圖係為本創作的視覺檢測裝置之參數設定方法之流程圖 第二圖係為本創作的自動優化參數的智能裝置之區塊示意圖。 第三圖係為本創作的檢測待測產品之示意圖。 The first picture is the flow chart of the parameter setting method of the visual inspection device created for this creation The second figure is a block diagram of the intelligent device that automatically optimizes parameters created by this author. The third picture is a schematic diagram of the product under test created by this author.
200:自動優化參數的智能裝置 200: Smart device that automatically optimizes parameters
210:待測產品 210: product to be tested
220:視覺檢測裝置 220: Visual inspection device
230:光學裝置 230: optical device
240:人工智慧裝置 240: Artificial Intelligence Device
242:參數設定器 242: Parameter Setter
244:真假瑕疵數量判斷器 244: True and False Defects Number Judge
246:處理器 246: processor
250:參數調整裝置 250: Parameter adjustment device
252:自動調整器 252: automatic adjuster
Claims (5)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW109212467U TWM606206U (en) | 2020-09-22 | 2020-09-22 | Intelligent device for optimizing parameters automatically |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW109212467U TWM606206U (en) | 2020-09-22 | 2020-09-22 | Intelligent device for optimizing parameters automatically |
Publications (1)
Publication Number | Publication Date |
---|---|
TWM606206U true TWM606206U (en) | 2021-01-01 |
Family
ID=75238219
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW109212467U TWM606206U (en) | 2020-09-22 | 2020-09-22 | Intelligent device for optimizing parameters automatically |
Country Status (1)
Country | Link |
---|---|
TW (1) | TWM606206U (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI737499B (en) * | 2020-09-22 | 2021-08-21 | 聯策科技股份有限公司 | A parameter setting method of a visual inspection system and smart system |
-
2020
- 2020-09-22 TW TW109212467U patent/TWM606206U/en unknown
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI737499B (en) * | 2020-09-22 | 2021-08-21 | 聯策科技股份有限公司 | A parameter setting method of a visual inspection system and smart system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11017259B2 (en) | Defect inspection method, defect inspection device and defect inspection system | |
US7822566B2 (en) | Method, device and program for setting a reference value for substrate inspection | |
US20190272628A1 (en) | Apparatus and method for enhancing optical feature of workpiece, method for enhancing optical feature of workpiece through deep learning, and non-transitory computer-readable recording medium | |
US5784484A (en) | Device for inspecting printed wiring boards at different resolutions | |
TWI715051B (en) | Machine learning method and automatic optical inspection device using the method thereof | |
Guo et al. | Research of the machine vision based PCB defect inspection system | |
TWI526683B (en) | Method of setting parameters for common setting between optical inspection units | |
TW201638881A (en) | Automatic optical inspection method of periodic patterns | |
TWM606206U (en) | Intelligent device for optimizing parameters automatically | |
TWI480541B (en) | Wafer pattern inspection apparatus | |
CN113379678A (en) | Circuit board detection method and device, electronic equipment and storage medium | |
TWI737499B (en) | A parameter setting method of a visual inspection system and smart system | |
WO2015015945A1 (en) | Defect candidate specification device, defect candidate specification method, defect determination device, defect inspection device, defect candidate specification program, and recording medium | |
JP2014055915A (en) | Appearance inspection device, appearance inspection method, and program | |
TW202117664A (en) | Optical inspection secondary image classification method which can effectively improve the accuracy of image recognition and classification | |
JP2768344B2 (en) | Pattern inspection apparatus and inspection method thereof | |
CN109407630B (en) | Parameter calculation method, device, terminal and readable storage medium | |
WO2000028309A1 (en) | Method for inspecting inferiority in shape | |
CN108508053B (en) | Method for detecting systematic infinitesimal physical defects | |
CN114199897A (en) | Parameter setting method and intelligent system of visual inspection system | |
CN116124786A (en) | Metal watch shell surface detection system and method based on machine vision | |
TWI759733B (en) | Method and system of artificial intelligence automatic optical inspection | |
KR20230075148A (en) | Apparatus and method for performing quality inspection of a product surface | |
TW201632874A (en) | Automatic lighting method for optical inspection and the optical inspection machine table thereof | |
CN113933309A (en) | Method for retesting defects in blind hole by AOI machine |