Strip steel periodic signal detection system and method

Abstract

The application discloses a strip steel periodic signal detection system, comprising: a surface multi-angle imaging device arranged in a box and used for intercepting an image of surface deformation of the strip steel; a speed sensor used for measuring the speed and length of the strip steel; a distance measuring device used for detecting the distance between the surface multi-angle imaging device and the surface of the strip steel; an internal high-pressure unit, which makes the air pressure in the box higher than the air pressure outside the box; and a calculation unit, which analyzes the image, determines the periodicity of the image, and superimposes the image according to the period. The application discloses a strip steel periodic signal detection method. The application solves the problem that the existing method can only be confirmed by manual visual inspection.

Description

Technical Field

[0001] This invention relates to strip steel surface quality inspection technology, and more specifically, to a strip steel periodic signal detection system and method. Background Technology

[0002] As a long-process product, strip steel involves numerous rollers during production. Therefore, periodic signals inevitably exist on the strip steel surface along its length. Accurately locating these periodic signals can greatly facilitate defect detection. Currently, the methods for locating periodic signals are typically manual visual observation or confirmation using oilstone polishing combined with visual observation.

[0003] In existing patent applications, such as Chinese patent application number 201911285462.1, a method for rapidly detecting pressure roller marks on production offset printing plates is provided, including the following steps: (1) sampling the production plate material; (2) heating the sampled plate material to age it, and then cooling it to form the plate; (3) exposing the plate material to light corresponding to the light source to form uniform dot or solid images, and then rinsing it in a developing machine to detect the uniformity of the dot or solid images after plate making, as an indicator for evaluating whether there are pressure roller marks. The application of this method can quickly determine whether the production plate material will cause product defects due to the above problems during subsequent storage, whether the production drying process or drag force is appropriate, and whether the surface state of each front roller matches the strength of the coating itself, thereby providing technical support for production, making timely adjustments, and preventing product defects in the later stages. This case uses a physical detection method to heat the material, which is obviously an offline detection, while this case is an online detection, and it is impossible to pre-treat the material by heating. Summary of the Invention

[0004] In view of the above-mentioned defects in the prior art, the purpose of the present invention is to provide a strip periodic signal detection system and method to overcome the shortcomings of the existing method which can only be confirmed by manual visual inspection.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] On the one hand, a strip periodic signal detection system includes:

[0007] A multi-angle surface imaging device, housed inside the enclosure, is used to capture images of the deformation on the surface of the strip steel.

[0008] Speed ​​sensor, used to measure the speed and length of strip steel;

[0009] A ranging device is used to detect the distance between the surface multi-angle imaging device and the surface of the strip steel;

[0010] An internal high-pressure unit ensures that the air pressure inside the enclosure is higher than the air pressure outside the enclosure.

[0011] The calculation unit analyzes the image, determines its periodicity, and superimposes the images periodically.

[0012] Preferably, the surface multi-angle imaging device includes a camera, a first light source, a second light source, and a third light source;

[0013] The angle between the first light source, the camera, and the normal is 10°-10°; the angle between the second light source, the camera, and the normal is 30°-10°; and the angle between the third light source, the camera, and the normal is 50°-10°.

[0014] Preferably, the fine-tuning range of the first light source, the second light source, and the third light source is ±3°.

[0015] Preferably, the speed sensor is a laser velocimeter with a measurement accuracy of 0.1 mm.

[0016] On the other hand, a method for detecting periodic signals in strip steel involves arranging the strip steel periodic signal detection system above the strip steel to perform the following steps:

[0017] S1. Image the surface of the strip steel using the surface multi-angle imaging device;

[0018] S2. Continuously acquire image data of the strip surface;

[0019] S3. The acquired image data is analyzed by the computing unit to extract periodic features and determine the periodicity of the acquired image.

[0020] S4. After determining the starting point of the periodicity, the images are superimposed periodically to enhance the defect signal, thereby detecting defects on the surface of the strip steel.

[0021] Preferably, in step S2, at least 3N period lengths of image data are acquired, where N depends on the least common multiple of the circumference of the rollers used for generation.

[0022] Preferably, in step S3, the FFT algorithm is used to extract periodic features to determine the periodicity of the acquired image;

[0023] The Fourier algorithm is used to perform periodic texture calculation on each image to calculate the periodic signal.

[0024] The present invention provides a strip steel periodic signal detection system and method. The advantage of this method is that it can automatically calculate the roll mark using an algorithm, without needing a close correlation with the circumference of the rolls on-site. This reduces the occurrence of period inaccuracies caused by various factors such as roll wear, which could lead to roll mark detection failures. For example, period inaccuracies can cause the image signal to be overwhelmed by the normal signal during superposition. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the arrangement of the strip periodic signal detection system of the present invention;

[0026] Figure 2 This is a schematic diagram of the angles between each light source, camera, and normal in the strip periodic signal detection system of the present invention;

[0027] Figure 3 This is a schematic flowchart of an embodiment of the strip periodic signal detection method of the present invention. Detailed Implementation

[0028] To better understand the above-mentioned technical solutions of the present invention, the technical solutions of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0029] Combination Figure 1 As shown, the present invention provides a strip periodic signal detection system, comprising:

[0030] A multi-angle surface imaging device is installed inside the housing 1. Its purpose is to capture image data of minute deformations on the surface of the strip steel 100.

[0031] The speed sensor, using a laser velocimeter, is used to measure the speed and length of the strip steel 100, with a measurement accuracy of 0.1mm.

[0032] To ensure imaging accuracy, in addition to using a laser velocimeter along the length of the strip 100, a high-precision ranging device is also used in the height direction of the strip 100 to ensure that the distance is always kept at a fixed value of 0. That is, the system needs a high-speed servo device to ensure that the distance between the multi-angle imaging device and the surface of the strip 100 remains unchanged.

[0033] An internal high-pressure unit ensures that the air pressure inside the enclosure is at least 0.1 MPa higher than the air pressure outside the enclosure.

[0034] The computing unit analyzes the image, determines its periodicity, and then overlays the images periodically.

[0035] Combination Figure 2As shown, the multi-angle imaging device employs various illumination methods to image the surface of the steel strip 100, including a camera 2, a first light source 3, a second light source 4, and a third light source 5. The first light source 3, the second light source 4, and the third light source 5, together with the camera 2, form an imaging field. The angles they form with the normal are 10°-10°, 30°-10°, and 50°-10°, respectively. These angles are the relatively ideal imaging angle process parameters obtained from current testing, and can be finely adjusted within ±3 degrees.

[0036] The present invention also provides a method for detecting periodic signals in strip steel, wherein the strip steel periodic signal detection system of the present invention is arranged above the strip steel 100 to perform the following steps:

[0037] S1. Capture fine deformation image data of the surface of strip steel 100 using a multi-angle surface imaging device;

[0038] S2. Continuously acquire image data of the strip surface, acquiring at least 3N cycle lengths, where N depends on the least common multiple of the circumference of the rollers used for generation;

[0039] S3. Analyze the acquired strip image data and extract the texture information generated by the roller's rotation. Since the roller rotates, the texture generated by each rotation will inevitably produce periodicity when mapped onto the strip surface. Use FFT calculations to extract periodic features and determine the periodicity of the acquired images.

[0040] The data processing procedure is as follows: The acquired images were captured transiently from three different angles. The images were then processed using photometric stereo processing. Photometric stereo processing uses multiple images to reconstruct the three-dimensional structure of an object's surface. It requires that the relative positions of the object and the camera remain constant. Then, the object is illuminated by light sources from different directions, producing different brightness and darkness effects. Potential roller marks were obtained through photometric stereo processing. The Fourier algorithm was used to perform periodic texture calculations on each individual image, yielding a periodic signal.

[0041] S4. After calculating the periodicity in the image and determining the starting point of the periodicity, the images are superimposed according to the period. The signal in the defective region is amplified, while the signal in the background region without defects is not amplified after superposition due to the randomness of the background signal. The corresponding images are summed pixel by pixel and superimposed to generate a new image representation with multi-pixel superposition. When the signal is enhanced to a sufficiently significant level, the signal processing and segmentation capabilities can be increased.

[0042] Example

[0043] Combination Figure 3As shown, there are three types of roller circumferences on a cold rolling production line: L1, L2, and L3. The strip length collected by the system must be at least a common multiple of L1, L2, and L3. In order to ensure the accuracy of the calculation, it is necessary to cover at least three cycles. By analyzing the image information collected in (1), the periodicity T of the collected image sequence is calculated using the FFT algorithm and the photometric stereo method. The periodicity is used to divide the collected image signal into segments according to the cycle. By accumulating the image pixel by pixel, the defect signal is enhanced, and the defects existing on the strip can be detected.

[0044] Those skilled in the art should recognize that the above embodiments are merely illustrative of the present invention and are not intended to limit the present invention. Any variations or modifications to the above embodiments that are within the spirit and essence of the present invention will fall within the scope of the claims of the present invention.

Claims

1. A strip steel periodic signal detection system, characterized in that, include: A multi-angle surface imaging device, housed inside the enclosure, is used to capture images of the deformation on the surface of the strip steel. Speed ​​sensor, used to measure the speed and length of strip steel; A ranging device is used to detect the distance between the surface multi-angle imaging device and the surface of the strip steel; An internal high-pressure unit ensures that the air pressure inside the enclosure is higher than the air pressure outside the enclosure. The computing unit analyzes the image, determines its periodicity, and then superimposes the images periodically. The surface multi-angle imaging device includes a camera, a first light source, a second light source, and a third light source; The angle between the first light source, the camera, and the normal is 10°-10°; the angle between the second light source, the camera, and the normal is 30°-10°; and the angle between the third light source, the camera, and the normal is 50°-10°.

2. The strip periodic signal detection system according to claim 1, characterized in that: The fine-tuning range of the first light source, the second light source, and the third light source is ±3°.

3. The strip periodic signal detection system according to claim 1, characterized in that: The speed sensor is a laser velocimeter with a measurement accuracy of 0.1 mm.

4. A method for detecting periodic signals in strip steel, characterized in that, A strip periodic signal detection system as described in any one of claims 1-3 is arranged above the strip to perform the following steps: S1. Image the surface of the strip steel using the surface multi-angle imaging device; S2. Continuously acquire image data of the strip surface; S3. The acquired image data is analyzed by the computing unit to extract periodic features and determine the periodicity of the acquired image. S4. After determining the starting point of the periodicity, the images are superimposed periodically to enhance the defect signal, thereby detecting defects on the surface of the strip steel.

5. The strip periodic signal detection method according to claim 4, characterized in that: In step S2, at least 3N period lengths of image data are collected, where N depends on the least common multiple of the circumference of the rollers used for generation.

6. The strip periodic signal detection method according to claim 5, characterized in that: In step S3, the FFT algorithm is used to extract periodic features to determine the periodicity of the acquired image; The Fourier algorithm is used to perform periodic texture calculation on each image to calculate the periodic signal.

Images