TW201413258A - Central control type automatic recognition and marking method of defective articles - Google Patents

Abstract

This invention relates to a central control type automatic recognition and marking method of defective articles, in which a plurality of units are used to perform image analysis on the articles under inspection. When a computer of the units determines that the articles under inspection are suspicious defective articles, the surface image data thereof is sent to a central console and displayed on a screen for re-confirmation by operational personnel. When they are confirmed as the defective articles, the central console enables the unit identifying the defective articles to produce a defective article mark on the surface of the articles under inspection confirmed as the defective articles, followed by other inspection and recognition steps on the articles under inspection. Therefore, one operator can monitor a plurality of units at the same time without having to walk back and forth among the units, thereby allowing automated operation of a plurality articles under inspection so as to provide a labor-saving efficient operation method.

Description

Central control type defective product automatic identification marking method

The invention relates to a central control automatic identification marking method, in particular to a method for an operator to simultaneously monitor multiple sets of detection markings in a central control mode for product shape detection, and automatically generate markings on the surface of the detected defective products.

In the manufacturing process, electronic components such as integrated circuit components (ICs) are used to ensure the quality of the shipment. After the integrated circuit components complete the assembly and print the product mark, etc., the functional function detection of the integrated circuit components is performed. In addition, the appearance shape is detected, and the detection of the appearance shape is used to determine whether the appearance shape of the integrated circuit component is dirty, notched, or the product mark is wrong.

In order to judge whether the shape and shape of the above-mentioned integrated circuit components are in good condition, and mark them on defective products for subsequent repair work, most of the industry currently uses a single-machine type inspection machine to perform computer-aided appearance inspection with manual mode on the defective product. Or the method of attaching the defective product mark, taking the tape-type integrated circuit component for detection and the manual defective product marking operation as an example, mainly for conveying the take-up type integrated circuit component to be detected to the image of the detecting machine Position, the image data of the surface of the integrated circuit component to be detected is transmitted from the image capturing device above the image capturing position to the computer, and the computer analyzes whether the shape of the integrated circuit component is in good condition by the image identification means, when the computer analyzes and judges The detecting integrated circuit component is a suspected defective product, and the image of the detected integrated circuit component is displayed on the screen, and a warning message is sent. At this time, the operator walks to the detecting machine and confirms whether it is not on the screen. Good product, After confirming that it is a defective product, the operator manually presses the switch to first move the integrated circuit component detected as a defective product in the tape-type integrated circuit component from the image capturing position to the marking working position, and then The operator manually marks or sticks the surface of the integrated circuit component of the defective product to generate a defective product mark, and then manually presses the switch to return the integrated circuit component to the image capturing position, and then continues to perform the next integrated circuit to be detected. Component inspection work.

However, the use of the aforementioned detecting machine can provide a function of automatically conveying the object to be inspected and image-assisted viewing, but the detecting machine is independent of each machine, and therefore, each operator can only manage 3 to 4 parts of the test. Only the machine, but the labor cost is too high, and because the detector finds a suspected defective product, confirms whether it is a defective product or a defective product marking operation, the operator must go to the testing machine and then manually write or paste. In the operation of defective product marking, the operator cannot directly manipulate the inspection machine at the remote end to perform automated inspection and marking operations, and there is a problem that the marking operation is inefficient.

The main object of the present invention is to provide a method for automatically identifying and marking a centrally controlled defective product, and to solve the problem that an existing operator only supervises 3 to 4 detection marks for performing defective product identification and manual marking operations, which requires excessive labor. .

In order to achieve the foregoing object, the method for automatically identifying the central control type defective product proposed by the present invention comprises: respectively feeding the plurality of detected objects to be detected to the image capturing position of the plurality of single machines; and each single machine uses the image capturing device to obtain the moving position. The surface image data of the test object at the image capturing position is collected by the computer of each single machine and taken from the image picker The surface image data of the test object obtained by the computer of each single machine is analyzed by image identification means whether the surface image data of the detected object conforms to the correct image data of the surface of the test object built in the computer of the single machine, wherein: when analyzing the surface image data of the test object When the correct image data is met, the computer that returns to each stand-alone computer analyzes whether the collected surface image data of the test object meets the correct image data of the surface of the test object built in the computer of the stand-alone computer by image recognition means; when analyzing the surface image data of the test object, When the device is in compliance with the correct image data of the test object and is judged to be a defective product, the following steps are performed: the surface image data of the test object determined to be a defective product is sent to the central control center computer by the computer of the single machine; central control The central control computer of the station collects the surface image data of the suspected defective product detected by the computer of the single machine, and presents the surface image data of the detection object of the suspected defective product on the screen of the central console; the operator displays according to the screen Suspected defective product surface image to confirm whether it is a defective product And input a message that is confirmed as a good product or a message that is confirmed as a defective product, wherein: when the computer of the stand-alone computer collects the information that the operator inputs the suspected defective product detection as a good product, the machine returns to the single machine. The computer analyzes whether the collected surface image data of the test object conforms to the correct image data of the surface of the test object built in the computer of the stand-alone computer by means of image identification; when the computer of the stand-alone computer collects the operator from the central control computer and inputs the suspected defective product test object When the message of the defective product is confirmed, the computer device of the stand-alone unit causes the defective device to produce a defective product mark on the surface of the test object which has been confirmed as a defective product. Remember, and then return to the computer of the single machine to analyze whether the collected surface image data of the test object conforms to the correct image data of the surface of the test object built in the computer of the single computer by image identification means.

By the invention of the centrally-controlled automatic identification method for defective products, the main method is to adopt a central centralized management and control method, and the operator supervises a plurality of single machines at the same time to perform multiple detections simultaneously. When the computer in the stand-alone computer judges that the detected object is a suspected defective product, the operator can confirm at the center console whether the suspected defective product is not detected. The good product is confirmed as a defective product, so that the single machine automatically generates a defective product mark on the surface of the test object that has been confirmed as a defective product, and then carries on the automatic identification mark operation of the other test object, so that the central control type of the present invention is not The automatic identification method of good products allows an operator to supervise multiple single machines at the same time, and the operator does not need to move between the single machines, that is, most of the detection objects can be performed in an automated manner, thereby achieving a labor-saving and high-efficiency operation method.

As shown in FIG. 1, a flow chart of a preferred embodiment of the method for automatically identifying the central control type defective product of the present invention is disclosed. As can be seen from the flowchart, the central control type automatic product identification method mainly includes the following steps: sequentially feeding the plurality of detection objects to be detected to the image capturing positions of the plurality of single machines; The picker obtains the surface image data of the detecting object moved to the image capturing position, and collects the surface image data of the detecting object obtained from the image picker by each computer of the single machine; The computer-aided computer analyzes whether the collected surface image data of the test object conforms to the correct image data of the surface of the test object built in the computer of the stand-alone computer, wherein: when the surface image data of the test object meets the correct image data, the image is returned. Each single computer analyzes whether the collected surface image data of the test object conforms to the correct image data of the surface of the test object built in the computer of the stand-alone computer by image recognition means; when the surface image data of the test object does not conform to the correct image data of the test object surface, When it is determined that the product is suspected to be defective, the following steps are performed: the surface image data of the detection object that is determined to be a defective product is sent to the center console control computer by the computer of the single machine; the central control computer collects the single machine of the central console. The surface image data of the suspected defective product detected by the computer, and the surface image data of the suspected defective product is presented on the screen of the central console; the operator confirms the surface image of the suspected defective product according to the screen display Whether it is a defective product, and input a confirmation as a good product Or a message that is confirmed as a defective product, wherein: when the computer of the stand-alone computer collects the information that the operator inputs the suspected defective product detection as a good product, the computer that returns to the single machine is collected by the image identification means analysis. Whether the surface image data of the test object conforms to the correct image data of the surface of the test object built in the computer of the stand-alone computer; when the computer of the stand-alone computer collects the information of the suspected defective product test object from the central control computer and confirms that it is a defective product, The computer of the single machine causes the marking device to generate a defective product mark on the surface of the detected object which has been confirmed as a defective product, and then returns to the computer of the single machine to analyze the collected image by means of image identification means. Whether the surface image data of the test object conforms to the correct image data of the surface of the test object built in the computer of the single machine.

In the method for automatically identifying the central control type defective product of the present invention, when the surface image data of the detection object does not conform to the correct image data of the surface of the detection object, and the judgment is a suspected defective product, the detection object of the suspected defective product is stopped at the The image capturing position of the single machine, and in the above-mentioned operation, the operator confirms whether it is a defective product according to the surface image of the suspected defective product detected on the screen, and inputs a message confirming the good product or a message confirming the defective product: when the computer of the single machine The operator collects the information indicating that the suspected defective product is confirmed as a good product, and then moves the detected object; when the computer of the single machine collects from the central control computer, the operator inputs the suspected defective product and confirms that it is a defective product. In the case of the message, the pointing device generates a defective mark on the surface of the detected object which is confirmed as a defective product by the computer of the stand-alone computer, and then moves the detected object.

In the method for automatically identifying the central control type defective product of the present invention, when the surface image data of the detection object does not conform to the correct image data of the surface of the detection object, and the judgment is a suspected defective product, the detection object of the suspected defective product is stopped at the The image capturing position of the single machine, and in the foregoing, the operator confirms whether it is a defective product based on the surface image of the suspected defective product detected by the confirmation screen, and inputs a message confirming that it is a good product or a message confirming that it is a defective product, wherein: The computer of the stand-alone computer collects the information that the operator inputs the suspected defective product detection as a good product, moves the detected object, and returns to the computer of the single machine to analyze the collected surface of the detected object by image recognition means. Whether the image data conforms to the surface of the test object built into the computer of a single computer The step of correct image data; when the computer of the stand-alone computer collects the information indicating that the suspected defective product is confirmed as a defective product from the central control computer, the single device causes the detected product that has been confirmed as the defective product to move to a marking operation Positioning, by the marking device disposed at the marking working position, generating a defective product mark on the surface of the detecting object that has been confirmed as a defective product, and returning the detected product of the marked defective product to the image capturing position, and returning to the returning position The computer of the single machine analyzes whether the collected surface image data of the test object conforms to the correct image data of the surface of the test object built in the computer of the single computer by image identification means.

In the method for automatically identifying the central control type defective product of the present invention, when the surface image data of the detection object does not conform to the correct image data of the surface of the detection object and is judged to be a defective product, the detection object of the suspected defective product continues to be self-imaged. The position moves toward a marked working position, and the other steps of conveying the detected object to be detected to the image capturing position of the plurality of single machines are continued, and whether the defective surface is confirmed by the operator according to the surface image of the suspected defective product detected by the screen. And inputting a message confirming that it is a good product or a message confirming that it is a defective product, wherein: when the computer of the stand-alone computer collects the information from the central control computer and the operator inputs the suspected defective product detection as a good product, the test object is ordered. The step of marking the work position and returning to the computer of the single machine to analyze whether the collected surface image data of the test object conforms to the correct image data of the surface of the test object built in the computer of the single machine; when the computer of the stand-alone computer is controlled from the center The computer collects the operator to input the suspected defective product and confirms that When the yield of the post, so that the stand-alone computer of the marking means disposed on the working position of the mark is moved to the work position mark A defective product mark is generated on the surface of the test object that has been confirmed as a defective product, and the detected product of the marked defective product is sent out through the marked work position, and the computer that returns to the single machine analyzes the collected test by image identification means. Whether the surface image data conforms to the correct image data of the surface of the test object built in the computer of the single machine.

In the method for automatically identifying the mark of the central control type defective product of the present invention, the single machine system obtains the surface image data of the two test objects moved to the image capturing position by using two sets of image pickers, and collects the data from the computer of each single machine. The surface image data of the detection object obtained by the image capturing device and the computer identification method of each single computer respectively analyze whether the surface image data of the detection object obtained by the image capturing device is consistent with the built-in test object of the computer. Correct image data on the surface; when a single computer collects information from the central control computer and the operator inputs any of the suspected defective products, it is determined by the computer of the single machine to move to the marking work position. The surface of the test object that has been confirmed as a defective product generates a defective product mark, and then returns to the computer of the single machine to analyze whether the collected surface image data of the test object conforms to the surface of the built-in test object of the single computer. The steps of the image data.

When a single computer collects information from the central control computer and the operator inputs the suspected defective product to confirm that it is a defective product, the computer marking device of the single machine has confirmed the second position at a certain point in the marking operation position. A defective product mark is generated on the surface of the test object of the defective product, or when the computer of the stand-alone computer collects the information that the operator inputs the suspected defective product and confirms that it is a defective product, the computer of the single machine has been Indeed It is considered that the detection object of the defective product is moved to the position of the marking work, and the marking device is moved by the computer of the single machine to the surface of the detection object which has been confirmed as the defective product, and the defective product mark is first generated, and then moved to The second product has been identified as a defective product mark on the surface of another test object in the test product of the defective product, and the test object of the two marked defective product is sent out through the mark work position, and then returned to the computer of the single machine to image The identification means analyzes whether the collected surface image data of the test object conforms to the step of correct image data of the surface of the test object built in the computer of the single machine.

According to the description of the preferred embodiment of the central control type automatic product identification method, the present invention mainly adopts a technical scheme of central centralized management and control, and the operator supervises a plurality of single machines and performs complex detection at a central console. The object sequentially carries out the steps of conveying the detection object, image analysis of the surface shape of the detection object, and the like, and when the computer in the single machine determines that the detection object is a suspected defective product, the operator can confirm the suspected defective product at the center console. Whether the test object is a defective product, and it is confirmed as a defective product, so that the single machine automatically generates a defective product mark on the surface of the test object that has been confirmed as a defective product, and then carries on the automatic identification mark operation of the other test object, so that the test is performed. The invention discloses a method for automatically identifying and marking the central control type defective product, so that an operator can supervise a plurality of single machines at the same time, and the operator does not need to move between the single machines, that is, the majority of the detection objects can be performed by the automatic operation mode, and the high-efficiency operation method for saving manpower is achieved. .

1 is a flow chart of a preferred embodiment of a method for automatically identifying a centrally controlled defective product according to the present invention.

Claims (6)

A method for automatically identifying and marking a centrally controlled defective product includes: feeding a plurality of detected objects to be detected to an image capturing position of a plurality of single machines; each of the single machines uses an image capturing device to obtain a surface image of the detecting object moved to the image capturing position Information, and the surface image data of the test object obtained from the image picker is collected by each computer of the stand-alone computer; and the computer-aided computer analyzes whether the collected surface image data of the test object conforms to the built-in test object of the computer. Surface correct image data, wherein: when analyzing the surface image data of the test object to meet the correct image data, the computer that returns to each stand-alone computer analyzes whether the collected surface image data of the test object conforms to the surface of the test object built in the computer of the stand-alone computer by image identification means. The step of correct image data; when it is determined that the surface image data of the detection object does not conform to the correct image data of the surface of the detection object, and is judged to be a defective product, the following steps are performed: the surface of the detection object which is judged to be a defective product by the computer of the single machine Image data is sent to the central console to control the computer; central control The central control computer collects the surface image data of the suspected defective product detected by the computer of the single machine, and presents the surface image data of the suspected defective product on the screen of the central console; the operator displays the suspect according to the screen display The surface image of the defective product is confirmed as a defective product, and a message indicating that it is a good product or a message indicating that it is a defective product is input, wherein: when the computer of the stand-alone computer collects the operator from the central control computer, the operator inputs the suspect. When the defective product is confirmed as a good product message, the computer that returns to the single machine analyzes whether the collected surface image data of the detected object conforms to the correct image data of the surface of the test object built in the computer of the single computer by using image recognition means; When the computer of the stand-alone computer collects the information indicating that the suspected defective product is confirmed as a defective product, the computer of the stand-alone computer causes the marking device to generate a defective product mark on the surface of the detected object that has been confirmed as a defective product, and then The computer that returns to the single machine analyzes whether the collected surface image data of the test object conforms to the correct image data of the surface of the test object built in the computer of the single computer by image identification means. The central control type defect automatic identification marking method according to claim 1, wherein the suspected defective product is determined when the surface image data of the analysis object does not conform to the correct image data of the detection object and is judged to be a defective product. The detection object stops at the image capturing position of the single machine, and in the foregoing, the operator confirms whether it is a defective product according to the surface image of the suspected defective product detected by the screen, and inputs a message confirming that it is a good product or a message confirming that it is a defective product. Time: when the computer of the stand-alone computer collects the information that the operator inputs the suspected defective product detection as a good product, and then moves the detected object; when the computer of the single machine collects from the central control computer, the operator inputs the suspected defective product. When the detected substance is confirmed as a defective product, the single-machine computer-marking device generates a defective product mark on the surface of the detected object which is confirmed as a defective product at the image capturing position, and then moves the detected object. The central control type defect automatic identification marking method according to claim 1, wherein when the surface image data of the analysis object does not conform to the correct image data of the surface of the detection object, and the suspected defective product is determined, the suspected defect is caused. The detected object of the product is stopped at the image capturing position of the single machine, and the operator confirms whether it is a defective product according to the surface image of the suspected defective product detected by the operator on the screen, and inputs a message confirming that it is a good product or confirming that it is a defective product. The message is: when the computer of the stand-alone computer collects the information from the central control computer and the operator inputs the suspected defective product detection as a good product, then moves the detected object, and returns to the computer of the single machine to use image identification means Analyze whether the collected surface image data of the test object conforms to the correct image data of the surface of the test object built in the computer of the stand-alone computer; when the computer of the stand-alone computer collects the information from the central control computer, the operator inputs the suspected defective product and confirms that it is a defective product. When the unit detects that the detected product that has been confirmed as a defective product is moved to a marking work position, the marking device disposed at the marking working position generates a defective product mark on the surface of the detecting object that has been confirmed as a defective product, and then Returning the detected product of the marked defective product to the image capturing position, and returning to the computer of the single machine to take a picture Identification of means of detection of the analyte surface image data collected meets the correct step of detecting image data of the surface built of stand-alone computer. The central control type defect automatic identification marking method according to claim 1, wherein the suspected defective product is determined when the surface image data of the analysis object does not conform to the correct image data of the detection object and is judged to be a defective product. The detecting object continues to move from the image capturing position toward a marking working position, and the other step of continuously transmitting the detecting object to be detected to the image capturing position of the plurality of single machines, and detecting the surface of the suspected defective product according to the screen displayed by the operator If the image confirms whether it is a defective product and enters a message confirming it as a good product or a message confirming it as a defective product, among them: When the computer of the stand-alone computer collects the information that the operator inputs the suspected defective product detection as a good product, the test object is sent out through the marking work position, and the computer that returns to the single machine is analyzed by the image identification means. Whether the collected surface image data of the test object conforms to the correct image data of the surface of the test object built in the computer of the single machine; when the computer of the stand-alone computer collects the information from the central control computer and the operator inputs the suspected defective product to be confirmed as a defective product, The computer of the single machine causes the marking device disposed at the marking working position to generate a defective product mark on the surface of the detecting object that has been confirmed as a defective product moved to the marking working position, and then the detecting object of the marked defective product passes the marking The working position is sent out, and the computer that returns to the single machine analyzes whether the collected surface image data of the detected object conforms to the correct image data of the surface of the detecting object built in the computer of the single computer by image identification means. The central control type defect automatic identification marking method according to claim 1, wherein the single machine system uses two sets of image capturing devices to respectively obtain surface image data of two detecting objects moved to the image capturing position, and each of the single machines The computer collects the surface image data of the detection object obtained from the second image capturing device, and analyzes whether the surface image data of the detection object obtained by the two image capturing devices are matched by the computer of each single computer by means of image identification means. The built-in test object surface correct image data; when a single computer collects from the central control computer and the operator inputs any of the two suspected defective products to be confirmed as a defective product, the computer is marked by the stand-alone computer A defect mark is generated on the surface of the test object which has been confirmed as a defective product at the marked work position, and then the computer of the stand-alone machine is returned to analyze the collected surface image data of the test object by image recognition means. Whether it is in accordance with the steps of correct image data on the surface of the built-in computer of the stand-alone computer; when the computer of the stand-alone computer collects the information from the central control computer and the operator inputs the suspected defective product, it is confirmed as a defective product, the computer program of the single machine The marking device generates a defective product mark on the surface of the detection object that has been confirmed as a defective product to be moved to the fixed point at a certain point of the marking operation position, and then causes the detection object of the two marked defective product to be sent out through the marking working position. Then, the computer that returns to the single machine analyzes whether the collected surface image data of the test object conforms to the correct image data of the surface of the test object built in the computer of the single machine by using image recognition means. The central control type defect automatic identification marking method according to claim 1, wherein the single machine system uses two sets of image capturing devices to respectively obtain surface image data of two detecting objects moved to the image capturing position, and each of the single machines The computer collects the surface image data of the detection object obtained from the second image capturing device, and analyzes whether the surface image data of the detection object obtained by the two image capturing devices are respectively matched by the computer of each single computer. The built-in test object surface correct image data; when a single computer collects from the central control computer and the operator inputs any of the two suspected defective products to be confirmed as a defective product, the computer is marked by the stand-alone computer A defective product mark is generated on the surface of the test object which has been confirmed as a defective product at the marked work position, and then returned to the computer of the single machine to analyze whether the collected surface image data of the test object conforms to the computer built-in of the single machine by image identification means. The step of detecting the correct image data on the surface of the object; when the computer of the stand-alone computer is collected from the central control computer, the operator inputs the Second doubt When the defect-like product is confirmed as a defective product, the computer that has confirmed the defective product moves to the marking operation position by the computer of the single machine, and the single-machine computer first moves the marking device to the second The surface of the test object which has been confirmed as a defective product first produces a defective product mark, and then moves to the surface of the other test object which has been confirmed as a defective product to produce a defective product mark, and then the second The test object of the defective product is sent out through the marking work position, and then the computer that returns to the single machine analyzes whether the collected surface image data of the test object conforms to the correct image data of the surface of the test object built in the computer of the single computer by image identification means. .