Improved self-adaptive CFAR detection method for defects of worsted wool fabric

Abstract

The invention discloses an improved self-adaptive CFAR (Constant False Alarm Rate) detection method for defects of worsted wool fabric, and relates to the technical field of defect detection of worsted wool fabric. The invention relates to an improved self-adaptive CFAR (Constant False Alarm Rate) detection method for defects of worsted wool fabric. The method comprises the following steps: shooting a color image of the worsted wool fabric by adopting a CCD (Charge Coupled Device) camera; converting the acquired color image into a grayscale image to obtain a grayscale value of each pixel; traversing the target grayscale image pixel by pixel by using a sliding window type CFAR detector; wherein the sliding window comprises a front edge sliding window and a rear edge sliding window; the combination of the leading edge sliding window and the trailing edge sliding window is used as a reference unit. According to the scheme, through the double-sliding-window structure design and the mean value self-adaptive selection algorithm, the false detection problem of the light and shade junction area is accurately restrained, and the false detection rate is remarkably reduced; the technical bottleneck that the false detection rate is high due to the fact that the fabric is uneven is broken through, double optimization of the false detection rate and the detection rate is achieved in a complex industrial scene, and a reliable technical scheme is provided for intelligent quality inspection of the worsted fabric.

Description

Technical Field

[0001] The invention relates to the technical field of worsted woolen fabric defect detection, and in particular to an improved worsted woolen fabric defect self-adaptive CFAR detection method. Background Art

[0002] In the worsted wool fabric production process, defect detection is a critical step in ensuring product quality, occurring throughout the entire process, from rough inspection to final inspection. Fabric defects are typically detected visually, followed by raw and finished fabric repairs before the finished product is stored. However, this method has significant limitations: high labor costs, low detection efficiency, and a high rate of missed detections due to subjective factors, making it difficult to meet the quality control requirements of large-scale production.

[0003] With the development of computer vision technology, algorithms such as histogram analysis, edge detection, threshold segmentation, Gabor transform, and wavelet transform have been applied to textile defect detection. While these methods have achieved a certain degree of automation, they still face challenges such as insufficient detection accuracy and high false positive rates, especially when detecting complex textures or subtle defects.

[0004] The recent rise of convolutional neural networks and deep learning technologies, with their powerful feature extraction capabilities, has made progress in defect detection. However, these methods have significant drawbacks: their performance is highly dependent on large, annotated datasets, making it prone to missing small defects. Their complex algorithmic structures and large number of parameters lead to high hardware deployment costs and poor real-time performance, limiting their widespread application in industrial scenarios.

[0005] Signal detection theory originated from radar target detection technology during World War II. Constant False Alarm Rate (CFAR) detection technology has matured and is widely used for target detection in time-varying, non-stationary, and non-Gaussian clutter environments. CFAR technology is known for its simple algorithm, adaptive threshold adjustment, high detection accuracy, and strong real-time performance.

[0006] The core detection principle of CFAR technology is similar to that of worsted fabric defect detection: both identify abnormal targets from complex backgrounds and separate abnormal features from background signals. Therefore, the inventors attempted to apply radar target CFAR detection technology to woolen fabric defect detection. However, the inventors' long-term practical research found that: Using CFAR technology for defect detection on worsted wool fabrics presents technical challenges: the unevenness of wool fabrics creates uneven light and dark areas in grayscale images. This phenomenon leads to excessive false detections at the boundaries between these areas (clutter edges). Therefore, solving these technical challenges is an urgent need for those skilled in the art.

[0007] The information disclosed in this background technology section is only intended to enhance understanding of the overall background of the invention and should not be regarded as an admission or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the Invention

[0008] In response to the above technical problems, an embodiment of the present invention provides an improved worsted fabric defect adaptive CFAR detection method to solve the problems raised in the above background technology.

[0009] The present invention provides the following technical solution: an improved worsted fabric defect adaptive CFAR detection method, comprising the following steps: A CCD camera is used to capture color images of worsted fabrics; Convert the collected color image into a grayscale image to obtain the grayscale value of each pixel; A sliding window CFAR detector is used to traverse the target grayscale image pixel by pixel; The sliding window includes a leading sliding window and a trailing sliding window; the combination of the leading sliding window and the trailing sliding window serves as a reference unit; Assuming that the grayscale value of worsted wool fabric follows Gaussian distribution, calculate the mean μ and variance σ of the wool fabric; The mean μ is replaced by the maximum value of the mean of the reference samples in the leading sliding window and the trailing sliding window, that is: ; Among them, the calculation formulas of μ1 and μ2 are: ; ; based on The variance σ of the woolen fabric is calculated based on the reference sample in the leading sliding window or the trailing sliding window; The formula for calculating variance σ is: ;or ; Set the decision thresholds T1 and T2; ; ; If the pixel to be detected x satisfies , it is judged as normal fabric; otherwise, it is judged as defective; Then, the detection window moves to the next pixel for detection until the entire image is detected; Where x is the pixel sample in the detection unit; is the pixel sample in the leading sliding window; is the pixel sample in the trailing sliding window; μ1 is the mean of the leading sliding window; μ2 is the mean of the trailing sliding window.

[0010] Preferably, the sliding window structure of the sliding window CFAR detector includes: a detection unit, a protection unit and a reference unit; The reference unit formed by the combination of the leading sliding window and the trailing sliding window is used to set the detection threshold; the protection unit is used to prevent larger defects from affecting the detection threshold.

[0011] Preferably, according to the preset It is required to determine t; where, is the false alarm probability (false detection rate).

[0012] Preferably, the calculation formula of t is: ; Among them, e is a natural constant.

[0013] Preferably, the detection method is suitable for automatic detection of points, lines, balls, holes, and other irregularly shaped defects in woolen fabrics.

[0014] An improved adaptive CFAR detection method for worsted fabric defects provided by an embodiment of the present invention has the following beneficial effects: the scheme of the present invention accurately suppresses the false detection problem in the light and dark boundary area through the double sliding window structure design and the mean adaptive selection algorithm, and significantly reduces the false detection rate; it breaks through the technical bottleneck of high false detection rate caused by uneven fabric, and realizes the dual optimization of false detection rate and detection rate in complex industrial scenarios, providing a reliable technical solution for the intelligent quality inspection of worsted fabrics. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 Schematic diagram of a sliding window for woolen fabric defect detection in the present invention; Figure 2 This is a flow chart of a specific implementation method of CFAR detection of woolen fabric defects in the present invention; in, Figure 2 The judgment criteria (5) and (7) in the patent refer to formula (5) and formula (7) in this patent. DETAILED DESCRIPTION

[0016] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0017] The following is combined with Figure 1-2 , and specific implementation methods are further described to illustrate the present invention.

[0018] In response to the problems mentioned in the above background technology, an embodiment of the present invention provides an improved worsted fabric defect adaptive CFAR detection method to solve the above technical problems. The specific technical solution is as follows.

[0019] 1. Design concept of an improved adaptive CFAR detection method for worsted fabric defects provided by the present invention Radar detection technology is primarily used for detecting targets such as aircraft and ships. Its detection principle is based on the difference in electromagnetic wave reflection between the target and the surrounding environment. The radar system receives the echo signal reflected by the target and compares it with a pre-set or dynamically calculated threshold. When the echo intensity exceeds the threshold, the target is determined to be present, achieving adaptive detection and accurately identifying and locating the target object.

[0020] Woolen fabric defects appear as irregular shapes, such as dots, lines, spheres, and holes. These defects appear bright or dark in grayscale images, distinguishing them from the grayscale of normal fabric. Therefore, the CFAR method can be used to detect woolen fabric defects based on the differences between the grayscale images of woolen fabric defects and those of normal fabrics.

[0021] The structure and decision criteria of the wool fabric defect detector are different from those of the radar target CFAR detector. The method proposed in this invention is applicable to the automatic detection of irregular shaped defects such as points, lines, balls, and holes in worsted wool fabrics. The specific detection principle is as follows: (1) The present invention adopts sliding window detection technology to detect fabric defects. Figure 1 As shown; Figure 1 The red unit in the figure is the detection unit (pixel). The reference sliding window uses sub-sliding window technology and is divided into a leading sliding window and a trailing sliding window. The green part on the left is called the leading sliding window, and the yellow part on the right is called the trailing sliding window.

[0022] The cells in the leading and trailing sliding windows are called reference cells, which are used to set the detection threshold. The part between the detection cell and the reference cell is called the protection cell, which is used to prevent larger defects from affecting the detection threshold. represents the pixel samples in the leading sliding window, represents the pixel samples in the trailing sliding window, and x represents the pixel samples in the detection unit.

[0023] Assume that the grayscale of wool fabric pixels obeys Gaussian distribution, that is ; (1) In formula (1) Represents the average value of woolen fabrics, Represents the variance of the woolen fabric. The unevenness of the woolen fabric will cause uneven light and dark areas in the grayscale image. This phenomenon will cause excessive false detection rate at the junction of uneven light and dark (clutter edge). To solve this problem, we first estimate the mean of the reference samples in the leading and trailing sliding windows respectively. 1 and 2 namely: ; ; (2) Then, select the mean of the reference samples in the leading sliding window and the trailing sliding window 1 and The maximum value of 2 is used as the mean value of woolen fabric ,Right now: (3) The maximum sub-sliding window mean is used 1 or The reference sample in the leading or trailing sliding window at 2 estimates the variance of the woolen fabric ,Right now: ;or ; (4) (2) If the grayscale value x of the pixel sample in the detection unit meets the judgment criteria: or (5) If the pixel in the detection unit is defective, it is judged as normal fabric.

[0024] According to the false alarm probability (false detection rate) requirement P fa , the threshold factor t is determined by the following formula: ; (6) make ,but ; (7) From formula (7), we can see that the false positive rate Average value of woolen fabric and variance This means that the detector has a false detection rate of 0.01% for woolen fabrics with different textures, tones, and flatness. Can be adaptively maintained at a low level At the same time, it can detect fabric defects with the highest probability. In this way, this patented detection method can provide reliability and effectiveness for fabric defect detection and realize adaptive CFAR detection of fabric defects.

[0025] This patented method suppresses false detections caused by unevenness in woolen fabrics by employing a sub-window mean maximum selection (GO) technique. This ensures the reliability and effectiveness of fabric defect detection and enables adaptive CFAR detection of fabric defects. This method offers advantages such as a low false detection rate, a high defect detection rate, a simple algorithm, fast speed, and adaptability.

[0026] It's important to note that the texture, color, and smoothness of the worsted fabrics being inspected in production workshops vary greatly, and defects on these fabrics also vary in size and shape. Existing methods typically suffer from high false detection rates, high rates of missed detection of small objects, and an inability to adaptively adjust detection thresholds in such time-varying, non-uniform backgrounds. 2. Specific embodiments (1) Detection method The wool fabric image captured by the CCD camera is converted from color to grayscale to obtain a grayscale image, which is then sent to the sliding window CFAR detector.

[0028] The sliding window CFAR detector uses the reference samples in the leading sliding window or the trailing sliding window to estimate the mean μ and variance σ of the woolen fabric, and then calculates the error rate according to the set error rate. The threshold factor t is determined to give the decision thresholds T1 and T2; ; .

[0029] (2) Test results In order to verify the effectiveness of this solution in solving the core technical problem of high false detection rate in light and dark boundary areas caused by uneven fabrics, 100 images of worsted wool fabrics with significant light and dark uneven textures (such as samples with wrinkles and differences in weave density) were selected for special testing.

[0030] The image resolution is 2560×1440, and a total of 624 defects are annotated (including 368 defects in the edge area). The parameters of the sliding window CFAR detector are set as follows: reference unit width 3 pixels, protection unit width 10 pixels, false alarm probability = , and enable the leading / trailing edge sliding window mean selection mechanism.

[0031] (1) False detection control effect for the light-dark boundary area Single sliding window CFAR method (single sliding window mean): The average false positive rate of 100 fabrics with significant light and dark uneven textures is , mainly because the mean estimation is disturbed by the grayscale difference on both sides, resulting in threshold offset.

[0032] The solution of the present invention (double sliding window mean selection): the average false detection rate is only , which is reduced by 80.6% compared with the single sliding window CFAR method.

[0033] (2) Defect detection rate and adaptability of the overall detection rate: 615 out of 624 defects were successfully detected, with a detection rate of 98.56%, of which the detection rate of defects in the edge area was 97.82% (360 / 368), and the technical effect was significant.

[0034] In summary, the solution of the present invention accurately suppresses the false detection problem in the light-dark boundary area through the twin sliding window structure design and the mean adaptive selection algorithm, significantly reducing the false detection rate; at the same time, it significantly improves the defect detection rate in the edge area.

[0035] Experimental data show that the present invention has broken through the technical bottleneck of high false detection rate caused by uneven fabrics, achieved dual optimization of false detection rate and detection rate in complex industrial scenarios, and provided a reliable technical solution for intelligent quality inspection of worsted wool fabrics.

[0036] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. An improved adaptive CFAR detection method for worsted fabric defects, characterized in that: The following steps are involved: A CCD camera is used to capture color images of worsted fabrics; Convert the collected color image into a grayscale image to obtain the grayscale value of each pixel; A sliding window CFAR detector is used to traverse the target grayscale image pixel by pixel; The sliding window includes a leading sliding window and a trailing sliding window; the combination of the leading sliding window and the trailing sliding window serves as a reference unit; Assuming that the grayscale value of worsted wool fabric follows Gaussian distribution, calculate the mean μ and variance σ of the wool fabric; The mean μ is replaced by the maximum value of the mean of the reference samples in the leading sliding window and the trailing sliding window, that is: ; Among them, the calculation formulas of μ1 and μ2 are: ; ; based on The variance σ of the woolen fabric is calculated based on the reference sample in the leading sliding window or the trailing sliding window; The formula for calculating variance σ is: ;or ; Set the decision thresholds T1 and T2; ； ； If the pixel to be detected x satisfies , it is judged as normal fabric; otherwise, it is judged as defective; Then, the detection window moves to the next pixel for detection until the entire image is detected; Where x is the pixel sample in the detection unit; is the pixel sample in the leading sliding window; is the pixel sample in the trailing sliding window; μ1 is the mean of the leading sliding window; μ2 is the mean of the trailing sliding window.

2. The improved worsted fabric defect adaptive CFAR detection method according to claim 1, characterized in that: The sliding window structure of the sliding window CFAR detector includes: a detection unit, a protection unit and a reference unit; The reference unit formed by the combination of the leading sliding window and the trailing sliding window is used to set the detection threshold; the protection unit is used to prevent larger defects from affecting the detection threshold.

3. The improved worsted fabric defect adaptive CFAR detection method according to claim 1, characterized in that: According to the preset It is required to determine t; where, is the false alarm probability.

4. The improved worsted fabric defect adaptive CFAR detection method according to claim 2, characterized in that: The calculation formula for t is: ； Among them, e is a natural constant.

5. The improved worsted fabric defect adaptive CFAR detection method according to claim 1, characterized in that: The detection method is suitable for automatically detecting points, lines, balls, holes, and other irregular-shaped defects in woolen fabrics.

Images