TW522217B - Device and method for detecting external appearance - Google Patents

Abstract

The present invention provides a device and a method for detecting an external appearance. The device for detecting external appearance comprises: an image detection device for detecting the contour of an object under examination; a light source for providing light to the object under examination; and a light reinforcement device for reinforcing the light source. The method for detecting external appearance comprises: (a) providing an object under examination; (b) detecting the image of the object under examination; (c) calculating the average line of the edge of the image; (d) taking a plurality of measurement points on the edge of the image and calculating the length between the measurement point and the average line; (e) using the length between the measurement point and the average line to calculate the standard deviation of the average line; and (f) separately adding the standard deviation to the length between the measurement point and the average line to obtain a plurality of reinforcement values and using the reinforcement values to determine whether the external shape of the object under examination meets the specification.

Description

522217 V. Description of the invention (1) The present invention relates to an appearance inspection, and in particular, to a method for reliably removing appearance defects, a special inspection device and a detection method thereof. House'- passive components other than passive components (for example, MLCC) are tested in a test machine to test the electrical properties and are packaged in paper tapes for shipment. Electricity ::: The package itself does not have the ability to detect the appearance of the product. Therefore, / and i packs are often mixed into the package and shipped, and the products that cause the drop in the yield rate are also inspected. At present, most of them are manual. The sampling method is up to date, and the external manual sampling test must be performed in a ready-made suit. The formula can only be calculated by calculating the electrode and the body: the visual processing method to determine the purpose of the appearance inspection, but this method is not suitable for the following reasons: θ because of the following reasons

1. The limitation of area and size is more contaminated than the internal ones during appearance inspection; defects such as defects, breakage, and poor edges are the parts that are opposite to each other, such as the electrode shown in Figure 1. Therefore, Rugao is more concerned about and The calculation of the size will not be able to fully detect the defective product; simply use the area 2 · For the example shown in Figure 2, the damage of f: 'in the left power will still make the measured area still consistent with However, when there are too many measurement areas for product offset area, so | ΰ S is poor, because the area measurement area will not be wrong with the production port J without area 0 tm S 丨 extension σσ automatic tuning · 3 · If the length and width of each position of the electrode are taken as shown in Figure 3, the method required to set the X and γ directions in the electrode is not taken, and the length of each segment of the electrode, but this method: :: 于 则 = 图 自For action

22217

V. Description of the invention The part where the width of each segment in the S electrode has been deformed and indented, but in the middle, but in fact, the middle figure is another example of misjudgment + and should be regarded as defective; 4. The image obtained by the product itself is uneven. The normal image is shown in Figure 5. Therefore, the probability of block judgment is quite high, and it is rebellious. The purpose of the present invention is to provide a kind of appearance detection device and its detection method. ί t observe bad products, so that the yield rate of shipment: big mention: remove the image detection device of the appearance of the test object; use ^. to read / original, and to strengthen the light source to strengthen the light source device. In the present invention, the light first arrives at the object to be measured. In the present invention, the light source is set relative to the image, and the light-reinforcing device and the d device are tilted. In addition, once the light source and the light source are set as mirrors, The semi-circular blouse image detection device is a charge coupled device. Qiao Qianyi, ^ In this Maoming, an appearance detection method is provided. · 枯 r 铖 (, an object to be tested; (b) detection to be included. (A) Amount: ηΛ take a plurality of quantities on the edge of the image Measure the point, and calculate whether the appearance of the object to be tested meets the specifications. ^ Use the reinforced value to judge Fang Bai ί ί :: Medium 'The appearance detection method further includes: (g) The level provided is to indicate the value of the pole, south丨 Tian—1bu-rm, 1 and 丨 W are used in the measurement point in the above step (d) to calculate ^ 2217 V. Description of the invention (3) __ Comparison of two quantitative direction ranges only on the same horizontal line , The length is 1 degree and the horizontal square 啖. The 3 detection methods to determine whether the inside of the object to be tested has a missing square 3 detection methods further include: ⑻Provide and use the measurement points in the above step ， to calculate the location and horizontal perimeter ^ The length between the two measuring points on the same-vertical line is compared with the length = to the standard range value, so as to determine whether the inside of the object to be tested is flawed or not. In the present invention, the appearance detection method further includes: Missing range value 'Use the image in step (b) above to detect the internal defect of the object to be tested' Defects within the range value is compared with the defect, thereby determining the analyte of defects is within specifications. 'Hereinafter, embodiments of the appearance detection device and the detection method of the present invention will be described with reference to the drawings. Type 81 is briefly explained. Figures 1 and 2 are schematic diagrams showing the appearance defects of passive components. Figures 3 and 4 are examples of misjudgments caused by conventional methods. Figure 5 is a diagram of misjudged by conventional devices. Fig. 6 shows a schematic diagram of the appearance detection device of the present invention; Fig. 7 shows an example of improving the image quality by the device of the present invention; and Figs. 8a to 8d show the appearance detection method of the present invention. schematic diagram.

522217

[Symbol description] 1 0 ~ Appearance detection device for passive components, 1〇 ~ Image detection device, 2〇 ~ Light source, 3 0 ~ Light reinforcement device, 31 ~ Reflector, 40 ~ Carrier device, 4 0 ~ 1 ~ Carrying section, 200 ~ passive components.

Refer to FIG. 6 'Appearance detection device of the present invention Device 10, light source 20, light-reinforcing farming device 3] 0 Including image debt measurement device! ◦ It can be used for debt measurement and placed in the bearing ==. ; Image debt " " :: 22〇 ° r ^ 1-^ Detecting the mounting element side, which is set in a manner relative to the image 20, with: w set, the light reinforcement device 30 to face the light source 31; With the above-mentioned Shadow Fool 7 light source 20, it can be the setting relationship between half-ring reflectors. The image quality can be improved by using, measuring, setting 10, light source 20, and light reinforcement device 30. Sound light and light reinforcement device 30 will light source 20 As shown in the reflection enhancement of the light, comparing it with the door edge of FIG. 5 'improved image quality is less prone to misjudgment as shown in FIG. 7 _ comparison', it is obvious that the price of the device according to the present invention. The following describes the appearance inspection of the passive component of the present invention with reference to Figure 8a.

522217

V. Description of the invention (5) Testing method 'It includes the following steps: The supporting device 40 of the carrier device 40 is set. The passive component 200 is first set on the passive component 200. The detection device 10 detects a line, such as the 8th "image 2 = the average line ^ point of the edge of the image, and calculates a number of measurements on the edge of the leveling image, each of which is 乂. The length of X χ) / (ί Calculated by the arrow in Figure 83 of the average line accuracy 2 (measured to the average line in ^ J 2 long ^ ^ ¥ Know the drought difference (the right side in Figure 83 is positive, Α Α on the left, / qi =, [(x. 2 + v + V +… + two choices, X · Production In order to strengthen the edge thick seam characteristics, the length of the measurement point to the average line should be plus the standard deviation ( That is, 丨 χ 'several enhancement values uw0, Xwi, Xw2 ... Xwn), and it is good. The value of ^ judges whether the appearance of the component meets the specifications, that is, whether the edge of the product is judged by the enhancement value Pi Er = Widow. Too rough or concave, convex too serious, and after calculating the edge defects, calculate the length value of χ and γ positions inside the product. First, provide a standard range value in the horizontal direction; Use the measurement points in the above steps to calculate the length between the measurement points located on the same horizontal line, and compare the length with the standard range value in the horizontal direction to determine whether the interior of the passive component has defects; then provide a standard with the water ^ direction The standard value of the vertical direction in which the measurement direction of the range value is perpendicular. Using the measurement points in the above steps, calculate the length between two measurement points on a vertical line perpendicular to the horizontal line, and mark the length and the vertical direction ^ The range value is compared to judge whether there is a defect inside the passive component. 'This' when there is an internal defect', its length must be smaller, such as the $ b solid

522217 V. Description of the invention (6) What it does not: Conversely, if the product is convex, its length value will inevitably be larger, as shown in Figure 8c. ’By this way, we can filter out defective products inside. Also ▲, you can use the power of Matrix Imaging Library (MIL 6 · 1) to find the narrow and short defects in the product, so that it can make up for the lack of judgment of the narrow and long sections; in detail, & Figure 8d : "For the second internal defect range value ', use the image in the above steps to detect this judgment: Dynamic: Defect, compare the internal defect range with the internal defect range value, and check whether the internal defect within 70 pieces meets the regulations. The product of the present invention is refined, and the method and method of shipment can indeed remove the appearance of bad products. The yield of ffiM is greatly improved. Although the present invention has been ^ limited to the date and time of the issue, any 孰: u Executive Example Exposure As above, however, it is not intended to be used within the scope of God. When an artist of item A can be considered without altering the scope of the present invention, he shall attach the modification and retouching of the attached application II. Therefore, the guarantee of the present invention is insulting. The scope of interest shall prevail.

Claims (1)

522217 Six patent application scopes: one type of appearance detection device, including dry: one = image detection device to detect the appearance of an object under test; a light source to provide light to the object under test; and a light reinforcement A device for supplementing the light source. > 2 · Appearance detection as described in item 1 of the patent application scope, wherein the light source is set in an inclined manner relative to the image detection device /, 3. · Appearance detection as described in the patent application scope The light-reinforcing device is a reflector ^ 4. The appearance detection device as described in item 3 of the scope of patent application, i The reflector is semi-circular. / 、 Y _ 5 · Appearance of the image detection device described in item i of the scope of the patent application 6. This image detection method is one of the appearance detection methods, including: (a) providing a test object; (b) detection The image of the object to be measured; (c) an average line out of the edge of the image; (d) taking a plurality of measurement points on the edge of the image, and calculating the length from the measurement point to the average line; e) Calculate the standard deviation using the lengths of the measurement points to the average line; and τ ^ (f) add the Zen difference to the lengths of the measurement points to the average line to obtain a plurality of enhancements And use these enhanced values to determine whether the appearance of a π > object meets specifications. ~ To be tested 7 · Appearance inspection method as described in item 6 of the scope of patent application, more 页 page 10 〇535-7132W; C90038; ITS.ptd 522217 (g) Provide a standard value in the horizontal direction, and calculate the two lengths on the same square using the above = such measurements and points, and compare the length with the value of the horizontal range. Check whether there is a defect inside the object. ".8. The appearance inspection method described in item 7 of the scope of patent application, further comprising: (h) Provide a vertical standard range value perpendicular to the measurement direction of the horizontal standard range value, and use the measurement points in step (d) above to calculate the The length between two measuring points is compared with the standard range value in the straight direction to determine whether there is a defect in the object under test. 9 · According to the appearance inspection method described in item 8 of the scope of the patent application, the package (i) provides an internal defect range value, and uses the image in the above step (b) to detect the internal defect of the object to be tested, The internal defect is compared with the value of the internal defect range, thereby determining whether the internal defect of the object to be tested meets the specifications.