Image detection scanning method and system for possible defects on object surface

Abstract

An image detection and scanning method for possible defects on the surface of an object and an image detection and scanning system thereof, the image detection and scanning method for possible defects on the surface of an object comprising: using a photosensitive element to capture a test image according to a test light, using a processor to determine whether the test image is normal, generating an alignment signal when the test image is normal, aligning the object to a detection position according to the alignment signal, after the object is aligned to the detection position, sequentially shifting one of a plurality of surface blocks divided along the extension direction of the surface of the object to the detection position, providing a detection light with a light source component facing the detection position, irradiating the detection position with a light incident angle less than or equal to 90 degrees relative to the forward normal of the surface block located at the detection position, and using a photosensitive element to capture detection images of each surface block sequentially located at the detection position according to the detection light after the object is aligned to the detection position.

Description

[Technical field]

[0001] This case is about an image detection and scanning method for possible defects on the surface of an object and an image detection and scanning system thereof. [Background technology]

[0002] Product defect detection is an extremely important part of the industrial production process. Defective products cannot be sold, or if defective intermediate products are sold to other manufacturers for processing, the final product will not work. One of the existing defect detection methods is to manually observe the product to be inspected with the naked eye or touch it with both hands to determine whether the product has defects, such as pits, scratches, color differences, defects, etc. However, the efficiency of manually detecting whether a product has defects is poor, and it is very easy to make a misjudgment, which will cause the problem of uncontrollable product yield. [Summary of the invention]

[0003] The present invention provides an image detection and scanning system for possible defects on the surface of an object, comprising a driving component, a light source component, a photosensitive element and a processor. The driving component carries an object, the surface of the object is divided into a plurality of surface blocks along an extension direction, the driving component rotates the object to align the object to a detection position, and after the object is aligned to the detection position, the plurality of surface blocks are sequentially displaced to the detection position. The light source component is arranged facing the detection position, providing a detection light to illuminate the detection position, and a light incident angle of the detection light is less than or equal to 90 degrees relative to a forward normal of the surface block located at the detection position. The photosensitive element is arranged facing the detection position, and captures a detection image of the surface blocks sequentially located at the detection position when the object is aligned to the detection position, and captures a test image according to a test light before the object is aligned to the detection position, wherein a photosensitive axis of the photosensitive element is parallel to the forward normal or between the forward normal and the extension direction. The processor is coupled to the photosensitive element to determine whether the test image is normal, and when the test image is normal, drives the driving component to align the object to the detection position.

[0004] The present invention provides an image detection and scanning method for possible defects on the surface of an object, comprising: using a photosensitive element to capture a test image according to a test light; using a processor to determine whether the test image is normal; generating an alignment signal when the test image is normal; aligning the object to a detection position according to the alignment signal; after the object is aligned to the detection position, sequentially displacing one of a plurality of surface blocks divided along an extension direction of the surface of the object to the detection position; providing a detection light with a light source component facing the detection position, wherein the detection light irradiates the detection position at a light incident angle less than or equal to 90 degrees relative to a forward normal of the surface block located at the detection position; and using the photosensitive element to capture detection images of each surface block sequentially located at the detection position according to the detection light after the object is aligned to the detection position, wherein a photosensitive axis of the photosensitive element is parallel to the forward normal or between the forward normal and the extension direction.

Brief Description of the Drawings

[0005] Figure 1 FIG. 4 is a schematic diagram of an embodiment of an image detection and scanning system for possible defects on the surface of an object according to the present invention.

[0006] Figure 2 for Figure 1 A functional block diagram of an embodiment of an image detection and scanning system for detecting possible defects on the surface of an object.

[0007] Figure 3 for Figure 1 A schematic diagram of an embodiment of the optical relative positions of an object, a light source assembly and a photosensitive element.

[0008] Figure 4 for Figure 1 A schematic diagram of another embodiment of the optical relative positions of an object, a light source assembly and a photosensitive element.

[0009] Figure 5 for Figure 1 A schematic diagram of another embodiment of the optical relative positions of an object, a light source assembly and a photosensitive element.

[0010] Figure 6 The flowchart is a method for scanning an image to detect possible defects on the surface of an object according to an embodiment of the present invention.

[0011] Figure 7 Corresponding to Figure 1 A schematic diagram of an embodiment of a predetermined surface defect of an object.

[0012] Figure 8 FIG. 2 is a schematic diagram of another embodiment of the image detection and scanning system for possible defects on the surface of an object of the present invention.

[0013] Fig. 9 for Figure 7 A schematic diagram of an embodiment of the optical relative positions of an object, a light source assembly and a photosensitive element.

[0014] Fig.10 for Figure 7 A schematic diagram of another embodiment of the optical relative positions of an object, a light source assembly and a photosensitive element.

[0015] Fig.11 The flowchart is another embodiment of the image detection and scanning method for possible defects on the surface of an object of the present invention.

[0016] Fig.12 It is a schematic diagram of an embodiment in which an object image includes sub-object images.

[0017] Fig.13 for Fig.12 A schematic diagram of an embodiment of an object image of FIG. [Specific implementation method]

[0018] Figure 1 FIG. 1 is a schematic diagram of an embodiment of an image detection and scanning system for possible defects on the surface of an object in the present case. Figure 1 The image detection scanning system for possible defects on the surface of an object is suitable for scanning the object 2 to obtain at least one detection image of the object 2. In some embodiments, the surface of the object 2 may have at least one surface defect, and the corresponding detection image will also present an image block of the surface defect. Here, the surface defect is a three-dimensional structure. Here, the three-dimensional structure is a millimeter (mm) size to a micrometer (μm) size.

[0019] The image detection and scanning system for possible defects on the surface of an object includes a driving component 11, a light source component 12 and a photosensitive element 13. Please refer to Figures 1 to 4 , Figure 2 for Figure 1 A functional block diagram of an embodiment of an image detection and scanning system for detecting possible defects on the surface of an object. Figure 3 and Figure 4 They are Figure 1 Schematic diagrams of two embodiments of the optical relative positions of the object 2, the light source assembly 12 and the photosensitive element 13. The light source assembly 12 and the photosensitive element 13 face a detection position 14 on the driving assembly 11. The driving assembly 11 carries the object 2 to be detected. The object 2 has a surface 21, and along an extension direction D1 of the surface 21 of the object 2, the surface 21 of the object 2 is divided into a plurality of surface blocks. Here, Figure 3 , 4 Taking the example that the surface 21 of the object 2 is divided into nine surface blocks and three of the surface blocks 21A- 21C are marked, the present invention is not limited thereto, and the surface 21 of the object 2 may also be divided into other numbers of blocks.

[0020] Please combine references Figures 1 to 6 , Figure 6 The flowchart of an embodiment of the image detection scanning method of the present invention for detecting possible defects on the surface of an object. The image detection scanning system for detecting possible defects on the surface of an object can execute a detection procedure. The driving component 11 sequentially shifts one of the plurality of surface blocks 21A-21C to the aforementioned detection position 14 in the detection procedure (step S01), and the light source component 12 emits a light beam L1 (hereinafter referred to as the detection light beam L1) toward each surface block 21A-21C located at the detection position 14 in the detection procedure (step S02), and as Figure 3 and Figure 4 As shown, the angle θ (hereinafter referred to as the light incident angle θ) between the incident direction of the detection light L1 and the forward normal 14A of each surface block 21A-21C located at the detection position 14 is greater than 0 degree and less than or equal to 90 degrees, that is, relative to the forward normal 14A, the detection light L1 irradiates the detection position 14 at a light incident angle θ greater than 0 degree and less than or equal to 90 degrees (step S03), so that the surface blocks 21A-21C are sequentially irradiated by the detection light L1 from the side or oblique direction at the detection position 14, and the surface blocks 21A-21C sequentially face the photosensitive element 13. In some embodiments, as Figure 3 , 4 As shown, the photosensitive axis 13A of the photosensitive element 13 is parallel to the positive normal line 14A; or Figure 5 As shown, the photosensitive axis 13A of the photosensitive element 13 is between the positive normal line 14A and the extension direction D1, that is, there is an angle α between the photosensitive axis 13A of the photosensitive element 13 and the positive normal line 14A. The photosensitive element 13 receives the diffuse light generated by the surface blocks 21A-21C receiving light, and the photosensitive element 13 captures the detection images of each surface block 21A-21C sequentially located at the detection position 14 according to the diffuse light (step S04).

[0021] For example, taking the case where the driving component 11 first displaces the surface block 21A to the detection position 14 in step S01, the detection light L1 provided by the light source component 12 irradiates the surface block 21A, and the photosensitive element 13 captures the detection image of the surface block 21A in step S04. The driving component 11 then displaces the object 2, and the driving component 11 further displaces the surface block 21B to the detection position 14 in step S01, and the photosensitive element 13 captures the detection image of the surface block 21B in step S04. The driving component 11 then displaces the object 2, and the driving component 11 further displaces the surface block 21C to the detection position 14 in step S01, and the photosensitive element 13 captures the detection image of the surface block 21C in step S04.

[0022] Therefore, according to the light incident angle θ greater than 0 degree and less than or equal to 90 degrees, that is, according to the sideways or obliquely incident detection light L1, if the surface 21 of the object 2 includes a groove-shaped or hole-shaped surface defect, the detection light L1 cannot irradiate the bottom of the surface defect, and the surface defect appears as a shadow in the detection image of the surface block 21A-21C. In this way, a detection image with a clear contrast between the surface 21 and the surface defect can be formed. The image detection scanning system or the inspector of the possible defects on the surface of the object can determine whether the surface 21 has a surface defect based on whether the detection image has a shadow (step S06). The inspector does not need to observe the object 2 with the naked eye or touch the object 2 with both hands to determine whether the surface of the object 2 has a surface defect. This will greatly improve the efficiency of surface defect detection and reduce the possibility of human misjudgment.

[0023] In one embodiment, the photosensitive element 13 can be implemented as a linear image sensor or a surface image sensor. Furthermore, the detection light L1 provided by the light source assembly 12 can be visible light, and the wavelength of the detection light L1 can be in the range of 380nm to 750nm, such as purple light with a wavelength of 380nm-450nm to red light with a wavelength of 620nm-750nm. Visible light can make surface defects on the surface 21 with a size of μm appear in the detection image.

[0024] In one embodiment, according to different light incident angles θ, surface defects with different depths appear with different brightness in the inspection image. Figure 4 As shown in FIG. 1 , when the light incident angle θ is equal to 90 degrees, the incident direction of the detection light L1 is perpendicular to the depth direction D2 of the surface defect. Regardless of the depth of the surface defect, the surface defect on the surface 21 does not generate reflected light and diffuse light because it is not irradiated by the detection light L1. The surface defects with deeper or shallower depths appear as shadows in the detection image, and the detection image has a poor contrast, or is close to no contrast. Figure 3 As shown, when the light incident angle θ is less than 90 degrees, the incident direction of the detection light L1 is not perpendicular to the depth direction D2 of the surface defect, and the detection light L1 irradiates a partial area of ​​the surface defect under the surface 21. The partial area of ​​the surface defect is irradiated by the detection light L1 to generate reflected light and diffused light. The photosensitive element 13 receives the reflected light and diffused light from the partial area of ​​the surface defect. The surface defect appears as an image with a brighter boundary (such as a convex defect) or a darker boundary (such as a concave defect) in the detection image, and the detection image has a better contrast.

[0025] Moreover, in the case of the same light incident angle θ less than 90 degrees, the photosensitive element 13 receives more reflected light and diffused light from the shallower surface defect than from the deeper surface defect, and the shallower surface defect presents a brighter image in the detection image compared to the surface defect with a larger depth-width ratio. Further, in the case of the light incident angle θ less than 90 degrees, if the light incident angle θ is smaller, the surface defect generates more reflected light and diffused light, the surface defect presents a brighter image in the detection image, and the brightness of the shallower surface defect in the detection image is also greater than the brightness of the deeper surface defect in the detection image. For example, the light incident angle θ of 30 degrees is less than the light incident angle θ of 60 degrees. Compared with the detection image corresponding to the light incident angle θ of 60 degrees, the surface defect presents a higher brightness in the detection image corresponding to the light incident angle θ of 30 degrees; and in the detection image corresponding to the light incident angle θ of 30 degrees, compared with the deeper surface defect, the shallower surface defect presents a higher brightness in the detection image.

[0026] Based on this, there is a negative correlation between the size of the light incident angle θ and the brightness of the surface defects on the inspection image. If the light incident angle θ is smaller, the shallower surface defects will appear brighter in the inspection image, that is, the image detection scanning system or inspector of the possible defects on the surface of the object will find it less likely to identify the shallower surface defects when the light incident angle θ is smaller, and the image detection scanning system or inspector of the possible defects on the surface of the object will find it easier to identify the deeper surface defects based on the darker images; on the contrary, if the light incident angle θ is larger, both shallower and deeper surface defects will appear darker in the inspection image, that is, the image detection scanning system or inspector of the possible defects on the surface of the object will be able to identify all surface defects when the light incident angle θ is larger.

[0027] Therefore, the image detection scanning system or the inspector of the possible defects on the surface of the object can set the corresponding light incident angle θ according to the preset hole depth d of the preset surface defect to be detected according to the aforementioned negative correlation, such as Figure 1 and Figure 2As shown, the image detection and scanning system for possible defects on the surface of an object further includes a light source adjustment component 16, which is coupled to the light source component 12. If it is desired to detect a deeper preset surface defect instead of a shallower preset surface defect, the light source adjustment component 16 can adjust the position of the light source component 12 according to the light incident angle θ calculated by the aforementioned negative correlation relationship and set a smaller light incident angle θ in step S03, and the light source adjustment component 16 drives the light source component 12 to output the detection light L1, so that the shallower preset surface defect appears as a brighter image in the detection image and the deeper preset surface defect appears as a darker image in the detection image; if it is desired to detect both the shallower and deeper preset surface defects, the light source adjustment component 16 can adjust the position of the light source component 12 according to the light incident angle θ calculated by the aforementioned negative correlation relationship and set a larger light incident angle θ (for example, 90 degrees) in step S03, and the light source adjustment component 16 drives the light source component 12 to output the detection light L1, so that both the shallower and deeper preset surface defects appear as shadows in the detection image.

[0028] For example, assuming that the object 2 is a spindle used in a safety belt assembly of an automobile, the surface defects mentioned above may be sand holes or air holes caused by dust or air during the manufacturing process of the object 2, or bump marks or scratches. Among them, the depth of sand holes or air holes is greater than that of bumps or scratches. If it is desired to detect whether the object 2 has sand holes or air holes but not to detect whether the object 2 has bumps or scratches, the light source adjustment component 16 can adjust the position of the light source component 12 according to the light incident angle θ calculated by the aforementioned negative correlation in step S03 and set a smaller light incident angle θ, so that the sand holes or air holes appear at a lower brightness in the detection image, while the bumps or scratches appear at a higher brightness in the detection image. The image detection scanning system or the inspector of possible defects on the surface of the object can quickly identify whether the object 2 has sand holes or air holes; if it is desired to detect whether the object 2 has bumps, scratches, sand holes and air holes, the light source adjustment component 16 can adjust the position of the light source component 12 according to the light incident angle θ calculated by the aforementioned negative correlation in step S03 and set a larger light incident angle θ, so that the bumps, scratches, sand holes and air holes all appear as shadows in the detection image.

[0029] In one embodiment, the light incident angle θ is related to a predetermined depth ratio of a predetermined surface defect to be detected. Figure 7As shown, taking the preset surface defect including the preset hole depth d and the preset hole radius r as an example, the preset hole radius r is the distance between any side surface in the preset surface defect and the positive normal 14A, and the ratio (r / d) between the preset hole radius r and the preset hole depth d is the aforementioned depth ratio (r / d), and the light incident angle θ is the arctangent (r / d). Based on this, the light source adjustment component 16 can adjust the position of the light source component 12 according to the depth ratio (r / d) of the preset surface defect to be detected in step S03 and set the critical angle of the light incident angle θ to the arctangent (r / d). The light incident angle θ must meet the conditions of being equal to or greater than the arctangent (r / d) and less than or equal to 90 degrees. After adjusting the position of the light source component 12, the light source adjustment component 16 drives the light source component 12 to output the detection light L1. In some embodiments, the preset hole radius r can be pre-set according to the size of the surface defect of the object 2 to be detected.

[0030] In one embodiment, if Figure 1 and Figure 2 As shown, the image detection and scanning system for possible defects on the surface of an object also includes a processor 15, which is coupled to a light source adjustment component 16. The processor 15 can calculate the light incident angle θ based on the aforementioned negative correlation and arctangent (r / d). The processor 15 then drives the light source adjustment component 16 to adjust the position of the light source component 12 based on the calculated light incident angle θ.

[0031] In one embodiment, if Figure 1 , Figure 3 and Figure 4 As shown, the object 2 has a cylindrical shape, and the surface 21 of the object 2 can be the side surface of the object 2, that is, the surface 21 is a cylindrical surface, and the surface 21 has a radian of 2π. Figure 1 and Figure 2 As shown, the driving assembly 11 includes a bearing element 111 and a stepper motor 112 connected to the bearing element 111. The bearing element 111 carries the object 2. In step S01, the stepper motor 112 can rotate in a clockwise direction or a counterclockwise direction. The stepper motor 112 drives the bearing element 111 to rotate to drive the object 2 to rotate. For example, the surface 21 includes surface blocks 21A-21C. The stepper motor 112 can rotate 120 degrees at a time to move each surface block 21A-21C to the detection position 14. In another embodiment, as Figures 8 to 10As shown, the object 2 has a rectangular parallelepiped or a cube shape, and the surface 21 of the object 2 may be a non-curved surface with a curvature equal to or close to zero. Here, along an extension direction D1 of the surface 21 of the object 2, the surface 21 may be divided into surface blocks 21A-21C. In step S01, the stepper motor 112 may drive the supporting element 111 to move parallel to the aforementioned extension direction D1 perpendicular to the positive normal line 14A, so as to move each surface block 21A-21C to the detection position 14.

[0032] In some embodiments, the aforementioned extension direction D1 may be, for example, Figure 1 The object 2 is rotated in the counterclockwise direction or clockwise direction with the central axis of the object 2 as the rotation axis. Figure 8 The major axis direction of the object 2 is shown.

[0033] In one embodiment, if Figure 1 and Figure 8 As shown, the light source assembly 12 may include a light emitting element. In another embodiment, as shown in FIG. Figure 3 , 4 , 9, and 10, the light source assembly 12 may include two light emitting elements 121 and 122, and the two light emitting elements 121 and 122 are symmetrically arranged on opposite sides of the object 2 relative to the forward normal line 14A. The two light emitting elements 121 and 122 illuminate the detection position 14 respectively, and the surface 21 is illuminated by the symmetrical detection light L1 to generate symmetrical diffuse light. The photosensitive element 13 sequentially captures the detection images of each surface block 21A-21C located at the detection position 14 according to the symmetrical diffuse light, so that the imaging quality of the detection image can be improved. In some embodiments, the light emitting elements 121 and 122 can be implemented by one or more light emitting diodes (LEDs); in some embodiments, each light emitting element 121 and 122 can be implemented by a laser light source.

[0034] In summary, according to an embodiment of the image detection scanning system for possible defects on the surface of an object and the image detection scanning method for possible defects on the surface of an object, the image detection scanning system for possible defects on the surface of an object can generate a detection image of the surface of the object. The image detection scanning system for possible defects on the surface of an object or the inspector can determine whether the surface of the object has surface defects based on the detection image. The inspector does not need to observe the object with the naked eye or touch the object with both hands to determine whether the surface of the object has surface defects. This will greatly improve the efficiency of surface defect detection and reduce the possibility of human misjudgment. Furthermore, the light incident angle is related to the preset hole depth or preset depth ratio of the preset surface defect to be detected. The image detection scanning system for possible defects on the surface of an object or the inspector can adjust the light incident angle so that the surface defect presents a corresponding brightness in the detection image. The image detection scanning system for possible defects on the surface of an object or the inspector can also efficiently identify the surface defect to be detected based on the detection images of different brightness.

[0035] In one embodiment, the image detection and scanning system for possible defects on the surface of an object further includes a test procedure and an alignment procedure. The image detection and scanning system for possible defects on the surface of an object sequentially executes the test procedure, the alignment procedure and the aforementioned test procedure. The photosensitive element 13 captures a test image according to a test light in the test procedure. In the alignment procedure, the object 2 has a preset reference point, and the driving component 11 moves the object 2 to align the preset reference point of the object 2 to the detection position 14. After the preset reference point of the object 2 is aligned to the detection position 14, the driving component 11 uses the preset reference point of the object 2 as the starting position and sequentially moves each surface block 21A-21C to the detection position 14.

[0036] For details, please refer to Figure 1 , 2 , 8, 11, the processor 15 is coupled between the photosensitive element 13 and the driving component 11. In the test procedure, the photosensitive element 13 captures a test image (step S07), the processor 15 receives the test image captured by the photosensitive element 13, and the processor 15 determines whether the test image is normal (step S08), and decides whether to drive the driving component 11 to execute the alignment procedure accordingly. If the test image is normal, it means that the photosensitive element 13 can capture a normal test image in step S04 of the detection procedure. The processor 15 generates a positioning signal after determining that the test image is normal (the judgment result is "yes"). The driving component 11 receives the positioning signal from the processor 15. The driving component 11 displaces the object 2 according to the positioning signal to align the object 2 to the detection position 14 (step S09). After the object 2 is aligned to the detection position 14, the image detection scanning system for possible defects on the surface of the object enters the detection procedure. The driving component 11 sequentially displaces each surface block 21A-21C to the detection position 14 in the detection procedure (step S01). The light source component 12 provides the detection light L1 to irradiate the detection position 14 in the detection procedure (steps S02 and S03). The photosensitive element 13 sequentially captures the detection image of each surface block 21A-21C located at the detection position 14 (step S04). No further details are given here.

[0037] In one embodiment, if Figure 2 As shown, the processor 15 is coupled to the stepper motor 112 . In step S09 , the processor 15 may send a positioning signal to control the stepper motor 112 , so that the stepper motor 112 rotates to drive the supporting element 111 to move or rotate, so as to position the object 2 to the detection position 14 .

[0038] In one embodiment, after the photosensitive element 13 captures the detection images of each surface block 21A-21C, the processor 15 receives the detection images of the surface blocks 21A-21C from the photosensitive element 13, and the processor 15 splices the detection images of the surface blocks 21A-21C into an object image (step S10). The image detection scanning system or the inspector of possible defects on the object surface can determine whether the entire surface 21 of the object 2 contains surface defects based on the spliced ​​object image.

[0039] In one embodiment, referring to Figure 8 , the image detection scanning system for possible defects on the surface of an object can be provided with a single photosensitive element 13, and this photosensitive element 13 performs image capture of multiple surface blocks 21A to 21C to obtain multiple detection images corresponding to the surface blocks 21A to 21C respectively. Then, the processor 15 splices the detection images of the surface blocks 21A-21C into an object image. In some embodiments, if the object 2 is cylindrical, the single photosensitive element 13 can perform image capture of multiple surface blocks 21A to 21C on the middle section of the object 2 to obtain multiple detection images corresponding to the surface blocks 21A to 21C respectively, and then the processor 15 splices the detection images of the surface blocks 21A-21C into an object image 23, such as Fig.13 As shown, Fig.13 The detection images 231 , 232 , and 233 of three surface blocks 21A- 21C are also illustrated.

[0040] In one embodiment, if Figure 1 As shown, the image detection scanning system for possible defects on the surface of an object can be provided with a plurality of photosensitive elements 13, and these photosensitive elements 13 respectively capture the detection images of the surface of the object 2 at different sections of the detection position 14. For example, taking the object 2 as a spindle, as Figure 1 As shown, one end of the object 2 has a relatively narrow structure, and the number of the photosensitive elements 13 can be three. The processor 15 splices the object image of the object 2 according to the detection images captured by the three photosensitive elements 13, such as Fig.12 The object image includes a sub-object image 22 ( Fig.12 The sub-object image 23 ( Fig.12 The sub-object image 24 ( Fig.12 (the lower part of the object image in ).

[0041] In one embodiment, the processor 15 can automatically determine whether the surface 21 of the object 2 includes surface defects, whether the surface 21 has different textures, and whether the surface 21 has attachments such as paint or oil stains according to the object image generated after the splicing. That is, the processor 15 can automatically determine the different surface morphologies of the object 2 according to the object image. In detail, the processor 15 includes an artificial neural network system, and the artificial neural network system has a learning phase and a prediction phase. In the learning phase, the object image input into the artificial neural network system is a known surface morphology. After the object image of the known surface morphology is input, the artificial neural network system performs deep learning based on the known surface morphology and the surface morphology category of the known surface morphology (hereinafter referred to as the preset surface morphology category) to establish a prediction model (i.e., composed of a number of hidden layers connected in sequence, each hidden layer has one or more neurons, and each neuron performs a judgment item). In other words, in the learning phase, the object image of the known surface morphology is used to generate the judgment items of each neuron and / or adjust the weight of the connection between any two neurons, so that the prediction result of each object image (i.e., the output preset surface morphology category) conforms to its known surface morphology.

[0042] For example, the aforementioned surface defects may be sand holes or pores, bumps or scratches. The image blocks showing different surface morphologies may be image blocks with sand holes of different depths, or image blocks without sand holes but with bumps or scratches, or image blocks with different surface roughness, or image blocks without surface defects, or image blocks with different contrasts and surface defects of different depth ratios produced by irradiating the surface blocks 21A-21C with detection light L1 of different light wavelength values, or image blocks with attachments of different colors. Here, the artificial neural network system performs deep learning based on the aforementioned object images of various different surface morphologies to establish a prediction model for identifying various surface morphologies. In addition, the artificial neural network system can classify the object images with different surface morphologies and pre-generate different preset surface morphology categories. Therefore, in the prediction stage, after the object image is generated and input into the artificial neural network system, the artificial neural network system executes the aforementioned prediction model based on the spliced ​​object image to identify the object image showing the surface morphology of the object 2 in the object image (step S11), and the object image is classified by the prediction model, that is, the artificial neural network system classifies the object image of the surface morphology of the object 2 according to a number of preset surface morphology categories (step S12). Among them, at the output end, the object image is predicted according to the preset surface defect category, that is, the percentage that it may fall into each category.

[0043] For example, taking the surface blocks 21A-21C as an example, the artificial neural network system executes the aforementioned prediction model according to the object image of the spliced ​​surface blocks 21A-21C. The artificial neural network system can use the object image of the object 2 to identify that the surface block 21A includes sand holes and bumps, the surface block 21B has no surface defects, the surface block 21C includes sand holes and paint, and the surface roughness of the surface block 21A is greater than the surface roughness of the surface block 21C; then, the preset surface shape categories include sand holes or pores, scratches or bumps, high roughness, low roughness, Taking the six categories of having attachments and not having surface defects as examples, the artificial neural network system can classify the detection image of the surface block 21A into the preset categories of sand holes or pores and scratches or bumps, and classify the detection image of the surface block 21B into the preset category of not having surface defects, and classify the detection image of the surface block 21C into the preset category of sand holes or pores and the preset category of having attachments, and classify the detection image of the surface block 21A into the preset category of high roughness, and classify the detection images of the surface blocks 21B and 21C into the preset category of low roughness. Based on this, the detection efficiency will be greatly improved by using the artificial neural network system to identify different surface morphologies, and the chance of manual misjudgment can be reduced.

[0044] In one embodiment, the deep learning performed by the aforementioned artificial neural network system may be implemented by a convolutional neural network (CNN) algorithm, but the present invention is not limited thereto.

[0045] In one embodiment, the photosensitive element 13 is the aforementioned linear image sensor, and the linear image sensor has a field of vision (FOV) close to 0 degrees. After the processor 15 receives the detection images of the surface blocks 21A-21C captured by the photosensitive element 13, in step S10, the processor 15 does not need to perform image processing based on the detection images of the surface blocks 21A-21C. The processor 15 can directly splice the detection images of each surface block 21A-21C into an object image.

[0046] In another embodiment, the photosensitive element 13 is the aforementioned area image sensor, which has a field of view of about 5 degrees to 30 degrees. In the detection images of the surface blocks 21A-21C captured by the photosensitive element 13, relative to the short side of each detection image, the middle area of ​​each detection image has a better imaging quality than other areas outside the middle area. Therefore, after receiving the detection images of the surface blocks 21A-21C captured by the photosensitive element 13, the processor 15 captures the middle area of ​​each detection image based on the short side of each detection image in step S10. In detail, in step S10, the processor 15 can capture the middle area of ​​each detection image according to a preset viewing angle within the field of view of the area image sensor. For example, the preset viewing angle can be 1 degree. The processor 15 captures the middle area of ​​each detection image corresponding to the preset viewing angle of 1 degree, such as a middle area with a width of one pixel. The processor 15 then splices the middle areas of each detection image into an object image to avoid incorporating other areas with poor imaging quality in the detection image into the object image, thereby improving the accuracy of the artificial neural network system in identifying the surface morphology of the object 2.

[0047] In one embodiment, in step S07, the photosensitive element 13 can capture an image of any surface block 21A-21C as a test image, and the processor 15 can compare the average brightness of the test image to determine whether it meets a preset brightness in step S08 to determine whether the test image is normal. If the average brightness of the test image does not meet the preset brightness (the judgment result is "no"), it means that the test image is abnormal, that is, the light incident angle θ set by the light source adjustment component 16 cannot correctly reflect the preset surface type to be detected. At this time, the image detection scanning system for possible defects on the surface of the object executes a calibration procedure (step S05), and the processor 15 controls the light source adjustment component 16 to readjust the position of the light source component 12 according to the aforementioned negative correlation or arctangent (r / d) in the calibration procedure to reset the light incident angle θ. After re-adjusting the position of the light source assembly 12, the light source adjustment component 16 drives the light source assembly 12 to emit another test light with a different light incident angle θ, so that the photosensitive element 13 captures an image of any surface block 21A-21C according to the other test light (step S07) to generate another test image. The processor 15 can then compare whether the average brightness of the other test image meets the aforementioned preset brightness (step S08). If the average brightness of the other test image still does not meet the preset brightness (the judgment result is "no"), the processor 15 controls the light source adjustment component 16 to re-adjust the position of the light source assembly 12 and re-adjust the light incident angle θ (step S05) until the average brightness of the test image captured by the photosensitive element 13 meets the preset brightness. When the average brightness of the test image meets the preset brightness, the processor 15 determines in step S08 that the test image is normal (the judgment result is "yes"), and the image detection and scanning system for possible defects on the surface of the object then executes the subsequent step S09 to perform the aforementioned alignment procedure and detection procedure.

[0048] In one embodiment, in step S08, the processor 15 can also determine whether the setting parameters of the photosensitive element 13 are normal according to whether the test image is normal. If the test image is normal (the judgment result is "yes"), it means that the setting parameters of the photosensitive element 13 are normal, and the processor 15 starts to generate the aforementioned alignment signal, so that the image detection scanning system for possible defects on the surface of the object enters the aforementioned alignment program and detection program in sequence. If the test image is abnormal (the judgment result is "no"), it means that the setting parameters of the photosensitive element 13 are abnormal, and the processor 15 does not generate the aforementioned alignment signal. At this time, the processor 15 further determines that the photosensitive element 13 has performed the adjustment operation of its setting parameters (step S13). If the photosensitive element 13 has performed the adjustment operation of its setting parameters (the judgment result is "yes"), the processor 15 generates a warning signal indicating that the photosensitive element 13 is abnormal (step S14). If the photosensitive element 13 has not performed the calibration operation of its set parameters (the judgment result is "No"), the image detection scanning system for possible defects on the surface of the object enters the aforementioned calibration procedure (step S05), and the processor 15 drives the photosensitive element 13 to perform the calibration operation of its set parameters in the calibration procedure. After the photosensitive element 13 performs the calibration operation, the photosensitive element 13 captures another test image in step S07, and the processor 15 then determines whether the other test image captured after the photosensitive element 13 performs the calibration operation is normal (step S08). If the processor 15 determines that the other test image is still abnormal (the judgment result is "No"), the processor 15 will then determine in step S13 that the photosensitive element 13 has performed the calibration operation (the judgment result is "Yes"), and the processor 15 generates a warning signal indicating that the photosensitive element 13 is abnormal in step S14.

[0049] In one embodiment, the image detection and scanning system for possible defects on the surface of an object further includes an audio-visual display unit, the aforementioned warning signal may include an image, a sound, or an image and a sound, and the audio-visual display unit may display the aforementioned warning signal. Furthermore, the image detection and scanning system for possible defects on the surface of an object may also have a network function, and the processor 15 may send the aforementioned warning signal to the cloud for storage through the network function, or send the warning signal to other devices through the network function, so that the user of the cloud or other device is informed that the photosensitive element 13 is abnormal, and then the photosensitive element 13 is debugged.

[0050] In one embodiment, the setting parameters of the photosensitive element 13 include a sensitivity value, an exposure value, a focal length value, a contrast setting value, or any combination of the above items. The processor 15 determines whether the average brightness or contrast of the test image meets the preset brightness in step S08 to determine whether the above setting parameters are normal. For example, if the average brightness of the test image does not meet the preset brightness, it means that any one of the setting parameters of the photosensitive element 13 is wrong, causing the average brightness or contrast of the test image to not meet the preset brightness; if the average brightness or contrast of the test image meets the preset brightness, it means that each of the setting parameters of the photosensitive element 13 is correct.

[0051] In one embodiment, in step S05, the photosensitive element 13 automatically adjusts the setting parameters according to a parameter setting file in the calibration procedure. The parameter setting file stores the setting parameters of the photosensitive element 13. The inspector can update the parameter setting file so that the photosensitive element 13 automatically adjusts the setting parameters according to the updated parameter setting file in the calibration procedure to correct the erroneous setting parameters.

[0052] In summary, according to an embodiment of the image detection and scanning method of possible defects on the surface of an object and its image detection and scanning system, the artificial neural network system can establish a prediction model based on deep learning of various object images with different surface morphologies. The artificial neural network system can identify and classify the different surface morphologies of the object based on the prediction model, which will greatly improve the efficiency of object detection and reduce the chance of manual misjudgment. Furthermore, the photosensitive element can capture the detection images at the same position on each surface block based on the aligned object, and the artificial neural network system can establish a more accurate prediction model based on the detection images at the same position, further reducing the chance of misjudgment.

[0053] Although the present invention has been disclosed as above by way of embodiments, it is not intended to limit the present invention. Any person having ordinary knowledge in the technical field may make some changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be determined by the scope of the patent application attached hereto.

Claims

1. An image detection and scanning system for detecting possible defects on the surface of an object, suitable for detecting an object, It is characterized in that Include: a driving assembly, carrying the object, wherein the surface of the object is divided into a plurality of surface blocks along an extension direction, the driving assembly displaces the object to align the object to a detection position, and after the object is aligned to the detection position, sequentially displaces the plurality of surface blocks to the detection position; a light source assembly disposed facing the inspection position, providing an inspection light beam to illuminate the inspection position, wherein a light incident angle of the inspection light beam relative to a forward normal of the surface block located at the inspection position is less than or equal to 90 degrees and greater than or equal to an inverse tangent value of a predetermined depth ratio of a predetermined surface defect to be inspected; a photosensitive element disposed facing the detection position, capturing detection images of the surface blocks sequentially located at the detection position after the object is aligned to the detection position, and capturing a test image according to a test light before the object is aligned to the detection position, wherein a photosensitive axis of the photosensitive element is between the forward normal line and the extension direction, and an angle is formed between the photosensitive axis and the forward normal line, wherein a magnitude of the light incident angle and a brightness of a surface defect presented on the detection image have a negative correlation; A processor is coupled to the photosensitive element to determine whether the test image is normal. When the test image is normal, the driving component is driven to align the object to the detection position; wherein, the test image is normal by determining whether the average brightness of the test image meets a preset brightness. If the average brightness of the test image does not meet the preset brightness, it indicates that the test image is abnormal. Then, the processor controls the light source component to readjust the position of the light source component according to the negative correlation or the inverse tangent value of the preset depth ratio in the calibration procedure to re-determine the light incident angle until the test image is determined to be normal; The processor determines whether the setting parameters of the photosensitive element are normal based on whether the test image is normal. If the test image is normal, it means that the setting parameters of the photosensitive element are normal. If the test image is abnormal, it means that the setting parameters of the photosensitive element are abnormal. The processor further determines whether the photosensitive element has performed the adjustment operation of its setting parameters. If the photosensitive element has performed the adjustment operation of its setting parameters, the processor generates a warning signal indicating that the photosensitive element is abnormal; if the photosensitive element has not performed the adjustment operation of its setting parameters, the adjustment program is entered, and the processor drives the photosensitive element to perform the adjustment operation of its setting parameters in the adjustment program; after the photosensitive element performs the adjustment operation, the photosensitive element captures another test image, and the processor determines whether the other test image captured after the photosensitive element performs the adjustment operation is normal.

2. The image detection and scanning system for possible defects on the surface of an object as claimed in claim 1, It is characterized in that The processor also receives the detection images of the plurality of surface blocks and splices the detection images of the plurality of surface blocks into an object image.

3. The image detection and scanning system for possible defects on the surface of an object as claimed in claim 2, It is characterized in that The processor captures the middle section area of ​​each detection image based on the short side of each detection image, and splices the plurality of middle sections into the object image.

4. The image detection and scanning system for possible defects on the surface of an object as claimed in claim 3, It is characterized in that The processor captures the middle area of ​​each detection image corresponding to the preset viewing angle based on a preset viewing angle within a field of view of the photosensitive element.

5. The image detection and scanning system for possible defects on the surface of an object as claimed in claim 2, It is characterized in that The processor comprises an artificial neural network system for identifying an image block showing the surface shape of the object in the object image according to a prediction model based on the object image.

6. The image detection and scanning system for possible defects on the surface of an object as claimed in claim 5, It is characterized in that The prediction model includes a plurality of preset surface morphology categories, and the artificial neural network system further classifies the image blocks of the surface morphology of the object according to the plurality of preset surface morphology categories.

7. An image detection and scanning method for possible defects on the surface of an object, suitable for detecting an object, It is characterized in that Include: Using a photosensitive element to capture a test image according to a test light; Using a processor to determine whether the test image is normal; When the test image is normal, a bit signal is generated; aligning the object to a detection position according to the alignment signal; After the object is aligned to the detection position, one of a plurality of surface blocks of the object divided along an extension direction of the surface thereof is sequentially displaced to the detection position; A light source assembly facing the inspection position provides an inspection light, wherein the inspection light irradiates the inspection position at a light incident angle less than or equal to 90 degrees and greater than or equal to an arctangent value of a predetermined aspect ratio of a predetermined surface defect to be inspected relative to a forward normal of the surface block located at the inspection position; After the object is aligned to the detection position, the photosensitive element captures detection images of each of the surface blocks sequentially located at the detection position according to the detection light, wherein a photosensitive axis of the photosensitive element is between the positive normal line and the extension direction, and an angle is formed between the photosensitive axis and the positive normal line; There is a negative correlation between the magnitude of the light incident angle and the brightness of a surface defect on the detection image; wherein, whether the test image is normal is determined by determining whether the average brightness of the test image meets a preset brightness; if the average brightness of the test image does not meet the preset brightness, indicating that the test image is abnormal, the processor controls the light source component to readjust the position of the light source component according to the negative correlation or the inverse tangent value of the preset depth ratio in the calibration procedure to re-determine the light incident angle until the test image is determined to be normal; The processor determines whether the setting parameters of the photosensitive element are normal based on whether the test image is normal. If the test image is normal, it means that the setting parameters of the photosensitive element are normal. If the test image is abnormal, it means that the setting parameters of the photosensitive element are abnormal. The processor further determines whether the photosensitive element has performed the adjustment operation of its setting parameters. If the photosensitive element has performed the adjustment operation of its setting parameters, the processor generates a warning signal indicating that the photosensitive element is abnormal; if the photosensitive element has not performed the adjustment operation of its setting parameters, the adjustment program is entered, and the processor drives the photosensitive element to perform the adjustment operation of its setting parameters in the adjustment program; after the photosensitive element performs the adjustment operation, the photosensitive element captures another test image, and the processor determines whether the other test image captured after the photosensitive element performs the adjustment operation is normal.

8. The image detection and scanning method for possible defects on the surface of an object as claimed in claim 7, It is characterized in that Also includes: Receiving the detection images of the plurality of surface blocks by the processor; The processor is used to stitch the detection images of the plurality of surface blocks into an object image.

9. The image detection and scanning method for possible defects on the surface of an object as claimed in claim 8, It is characterized in that In the step of splicing the detection images of the plurality of surface blocks into the object image by the processor, the processor captures the middle section area of ​​each detection image based on the short side of each detection image, and splices the plurality of middle sections into the object image.

10. The image detection and scanning method for possible defects on the surface of an object as claimed in claim 9, It is characterized in that In the step of splicing the detection images of the plurality of surface blocks into the object image by the processor, the processor captures the middle area of ​​each of the detection images based on a preset viewing angle within a field of view of the photosensitive element, and splices the plurality of middle areas into the object image.

11. The image scanning method for detecting possible defects on the surface of an object as claimed in claim 7, It is characterized in that Also includes: An artificial neural network system of the processor executes a prediction model according to the object image to identify an image block in the object image that presents the surface shape of the object.

12. The image detection and scanning method for possible defects on the surface of an object as claimed in claim 11, It is characterized in that In the step of executing the prediction model by the artificial neural network system, the artificial neural network system further classifies the image blocks of the surface morphology of the object according to a plurality of preset surface morphology categories.

Images