TW202331237A - Surface defect detection and early warning method for elongated products on production line capable of giving an immediate early warning to the personnel when there is a defect on product surface so as to fast determine possible defect positions - Google Patents

Abstract

Provided is a surface defect detection and early warning method for elongated products on a production line, which includes the following steps: (a) using a signal capturing module to capture surface information of an elongated product traveling on a conveyor belt; (b) determining whether a defect detection module detects a defect from the surface information; (c) in the case that the detection result of step (b) is affirmative, using a defect position estimation module to estimate the defect position of the defect on the surface of the elongated product; and (d) using an alarm module corresponding to the defect position to give an immediate early warning to the personnel of the defect position.

Description

Surface defect detection and early warning method for long-shaped products used in production line

The invention relates to a defect detection and early warning method, in particular to a surface defect detection and early warning method for long-shaped products on a production line.

During the forming process of the steel billet, various factors often cause defects on its surface that can affect the subsequent process products. Taking the small steel billet as an example, the surface of the small steel billet needs to be inspected by resident inspectors in the darkroom by means of magnetic particle inspection. However, the inspection method of magnetic particle inspection is not only labor-intensive and time-consuming, but also the inspection results are more likely to miss defects due to the reduction of the magnetic flux leakage effect caused by the large surface defects of the small billet. Therefore, it is necessary to find a solution.

Therefore, the object of the present invention is to provide a detection and early warning method for surface defects of elongated products on the production line.

Therefore, the present invention is used in the detection and early warning method of long-shaped product surface defects on the production line, including: (a) the signal acquisition module captures the surface information of the long-shaped product surface traveling on the conveyor belt; (b) defect detection Whether the module detects defects from the surface information; (c) if the detection result of the step (b) is yes, the defect position estimation module estimates the position of the defect on the surface of the elongated product Defect location; and (d) the alarm module corresponding to the defect location gives an immediate warning to the personnel of the defect location.

The effects of the present invention are: (1) when the product surface has a defect, the defect position estimation module can estimate the defect position of the defect on the surface of the elongated product, so that at least one warning module corresponding to the defect position can immediately Early warning to personnel, so that personnel can quickly find out the possible defect location; (2) Compared with the traditional magnetic particle flaw detection method, the present invention not only saves labor and time, but also does not have the magnetic particle flaw detection method caused by excessive defects on the surface of the small steel billet. The magnetic flux leakage effect is reduced and the defects are missed.

Referring to Figures 1 to 4, an embodiment of the system for implementing the method for detecting and warning the surface defects of long products on the production line of the present invention includes several light sources 11, several signal acquisition modules 12, and a partition 14 , a cleaning unit 15, a window 16, a storage module 17, a defect detection module 18, a defect position estimation module 19, and several warning modules 10 and other components. As shown in Fig. 2, in this embodiment, the elongated product may be, for example, a small steel billet 2, etc., but the present invention is not limited thereto.

As shown in FIGS. 2 and 3 , the light sources 11 are angularly spaced around the conveyor belt 3 , and the signal acquisition modules 12 are also angularly spaced around the conveyor belt 3 . Wherein, in this embodiment, each signal acquisition module 12 is used to acquire the surface information of the small steel billet 2 traveling on the conveyor belt 3, and may be, for example, a two-dimensional signal acquisition device , or a three-dimensional signal capture device, wherein the two-dimensional signal capture device can be a charge-coupled device (CCD), or a complementary metal oxide semiconductor (CMOS) camera, etc., and the three-dimensional signal capture device can be a pure laser scanning device, or a composite laser scanning device including a laser scanning device and a camera, etc.

As shown in Figure 3, in this embodiment, the partition 14, the cleaning unit 15 and the viewing window 16 are arranged below the small steel billet 2 and the conveyor belt 3, and are located at the bottom of the conveyor belt 3 and below. between light sources 11. Wherein, the viewing window 16 (such as a glass window, etc...) is arranged on the partition plate 14, so that the light projected by the light source 11 below can pass through the viewing window 16 and project to the small steel billet 2. In addition, the cleaning unit 15 is disposed above the partition 14 and opposite to it.

The storage module 17 is used to classify and store the sorted surface information, and can be, for example, a hard disk, a network attached storage (NAS), a server, or a workstation.

The defect detection module 18 uses feature analysis, artificial intelligence, deep learning algorithm, or other detection methods to detect whether there is a defect in the surface information.

The defect position estimation module 19 estimates the defect position on the surface of the steel billet 2 when the defect detection module 18 detects a defect in the surface information.

When the defect position estimation module 19 estimates at least one defect position on the surface of the steel billet 2, the at least one warning module 10 corresponding to the at least one defect position will warn the at least one defect position to the inspection personnel.

In this embodiment, each warning module 10 may be, for example, a warning light, etc. . . .

Referring to Figures 1, 3, 4, and 5, the embodiment of the method for detecting and early warning of surface defects of long products on the production line of the present invention includes the following steps. Firstly, as shown in steps S901, S902, and S903 in FIG. 5, pre-work is performed. Wherein, in step S901, the signal acquisition modules 12 and the light sources 11 are arranged around the conveyor belt 3 .

And, as shown in step S902 and FIG. 4, each small steel billet 2 is divided into several equal parts with equal lengths in advance. For example, in the present embodiment of FIG. In the length direction, each small steel billet 2 is divided into seven equal parts 21 to 27 of equal length. Moreover, as shown in step S903, it is necessary to set seven equal intervals at a predetermined interval D in the quality control detection area (as shown in FIG. 4 ) downstream of the signal acquisition modules 12 and the conveyor belt 3 . There are two warning modules 10 respectively corresponding to the equal parts 21 to 27 of each small steel billet 2, wherein, in this embodiment, the predetermined distance D may be, for example, about 2 meters.

Therefore, after the preceding operations of steps S901, S902, and S903 are completed, as shown in steps S91, FIG. 2, and FIG. The signal acquisition module 12 can acquire the surface information of the surface of the small steel billet 2 .

Next, as shown in step S92 , the surface information captured by the signal capture modules 12 is sorted and stored in the storage module 17 .

Next, as shown in step S93, does the defect detection module 18 detect defects from the surface information? If it is not detected, then as shown in step S94, no subsequent action is required.

On the contrary, if there is a detected defect, then as shown in step S95, the defect position estimation module 19 estimates the defect position. In this embodiment, the defect position estimation module 19 is based on the time when each equal portion 21 to 27 of the small steel billet 2 is captured by the signal acquisition module 12, and calculates the location of the defect at the Which equal parts of the steel billet 2 are on the surface, so the position of the defect on the surface of the steel billet 2 can be estimated.

Then, as shown in step S96, when the signal acquisition modules 12 have finished capturing the signals of all the equal parts 21 to 27 of the steel billet 2, and then the steel billet 2 is suspended in the position shown in FIG. In the quality control inspection area, the alarm module 10 corresponding to at least one of the segments 21 to 27 with defects will give an immediate warning of the defect location to personnel. For example, taking Fig. 4 as an example, because the defect position estimation module 19 estimates that the defect occurs in the equal section 23 and the equal section 25, the two warning modules corresponding to the equal section 23 and the equal section 25 10 (such as warning lights, etc...) will give an immediate warning, so inspectors can quickly find out possible defect locations through the warning module 10 whether the warning is issued or not.

To sum up, the method for detecting and early warning of surface defects of elongated products used in the production line of the present invention has at least the following advantages and effects: (1) When there is a defect on the surface of the elongated product, the defect position estimation module 19 can Estimate the defect position of the defect on the surface of the elongated product, so that at least one warning module 10 corresponding to the defect position can give an immediate warning to the personnel, so that the personnel can quickly find out the possible defect position; (2) the present invention is used in the production line Compared with the traditional magnetic particle flaw detection inspection method of the long-shaped product surface defect detection and early warning method, the present invention not only saves labor and time, but also does not reduce the magnetic flux leakage effect caused by the large surface defect of the small steel billet 2 in the magnetic particle flaw detection method. And produce the situation of missed judgment defect; Therefore can really reach the purpose of the present invention.

But the above-mentioned ones are only embodiments of the present invention, and should not limit the scope of the present invention. All simple equivalent changes and modifications made according to the patent scope of the present invention and the content of the patent specification are still within the scope of the present invention. Within the scope covered by the patent of the present invention.

1: system 10:Warning module 11: Light source 12: Signal acquisition module 14: Partition 15: Cleaning unit 16: Windows 17: Storage module 18: Defect detection module 19: Defect location estimation module 2: small billet 21~27: equal division 3: conveyor belt D: Predetermined spacing S901~S96: steps

Other features and effects of the present invention will be clearly presented in the implementation manner with reference to the drawings, wherein: Fig. 1 is a block diagram illustrating the implementation of the system of the embodiment of the present invention for the detection and early warning method of surface defects of elongated products on the production line; Fig. 2 is a schematic diagram viewed from one end of a small steel billet, illustrating that several signal acquisition modules and several light sources are arranged around the advancing small steel billet; Fig. 3 is a schematic diagram of a side view, illustrating that one of the light sources is located below the billet and the conveyor belt; Fig. 4 is a schematic diagram illustrating that in this embodiment, before capturing images of the billets, each billet must be divided into equal-length numbers along the length direction of each billet. and in the quality control inspection area, a plurality of warning modules respectively corresponding to the equal divisions of each small steel billet are arranged at equal intervals; and FIG. 5 is a flow chart illustrating the embodiment of the present invention for the detection and early warning method of surface defects of elongated products on the production line.

S901~S96: steps

Claims (8)

A detection and early warning method for surface defects of elongated products used on a production line, comprising the following steps: (a) The signal acquisition module acquires the surface information of the surface of the elongated product traveling on the conveyor belt; (b) Whether the defect detection module detects defects from the surface information; (c) if the detection result of the step (b) is yes, the defect position estimation module estimates the defect position of the defect on the surface of the elongated product; and (d) The warning module corresponding to the defect position gives an immediate warning to the personnel of the defect position. The surface defect detection and early warning method for long-shaped products on the production line according to claim 1, wherein, in the step (a), the signal acquisition module is a two-dimensional or three-dimensional signal acquisition device. The surface defect detection and early warning method for elongated products on the production line as described in claim 1, further includes the following step before the step (a): setting the signal acquisition module around the conveyor belt , and a light source, wherein the signal acquisition modules are angularly spaced around the conveyor belt, and the light sources are also angularly spaced around the conveyor belt. The surface defect detection and early warning method for elongated products on the production line as described in claim 1, wherein, in the step (b), the defect detection module uses feature analysis, artificial intelligence or deep learning calculations method to determine whether the defect exists in the surface information. The surface defect detection and early warning method for elongated products on the production line as described in claim 1, further includes the following steps between the step (a) and the step (b): the surface information is received After adjustment, the classification is stored in the storage module. The surface defect detection and early warning method for elongated products on the production line as described in claim 1, further includes the following steps before the step (a): along the length direction of each elongated product, each A long product is divided into several equal parts of equal length, and several equal intervals are set at a predetermined interval in the signal acquisition module and the quality control inspection area downstream of the conveyor belt. Corresponding to the warning module of the equal division of each elongated product; wherein, in the step (c), the defect position estimation module is based on each division of each elongated product According to the time when the signal is picked up by the signal pick-up module, calculate which equal parts of the surface of the long-shaped product the defect is on, so that the position of the defect on the surface of the long-shaped product can be estimated defect location. The surface defect detection and early warning method for elongated products on the production line as described in claim 6, wherein, in the step (d), when the signal acquisition module detects all the elongated products, etc. After the sub-signals are captured and the elongated product is stopped in the quality control inspection area, the alarm module corresponding to the sub-section of the defect alerts personnel of the defect location. The surface defect detection and early warning method for elongated products on the production line as described in Claim 1, wherein each warning module is a warning light.

Images