CN111539967B - Method and system for identifying and processing interference fringe region in terahertz imaging of focal plane - Google Patents

Method and system for identifying and processing interference fringe region in terahertz imaging of focal plane Download PDF

Info

Publication number
CN111539967B
CN111539967B CN202010333623.6A CN202010333623A CN111539967B CN 111539967 B CN111539967 B CN 111539967B CN 202010333623 A CN202010333623 A CN 202010333623A CN 111539967 B CN111539967 B CN 111539967B
Authority
CN
China
Prior art keywords
image
stripe
dark
focal plane
threshold
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202010333623.6A
Other languages
Chinese (zh)
Other versions
CN111539967A (en
Inventor
王军
张江威
刘琦
杨明亮
何美誉
谢哲远
张超毅
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
University of Electronic Science and Technology of China
Original Assignee
University of Electronic Science and Technology of China
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by University of Electronic Science and Technology of China filed Critical University of Electronic Science and Technology of China
Priority to CN202010333623.6A priority Critical patent/CN111539967B/en
Publication of CN111539967A publication Critical patent/CN111539967A/en
Application granted granted Critical
Publication of CN111539967B publication Critical patent/CN111539967B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/10Segmentation; Edge detection
    • G06T7/11Region-based segmentation
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00Image enhancement or restoration
    • G06T5/70Denoising; Smoothing
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/10Segmentation; Edge detection
    • G06T7/136Segmentation; Edge detection involving thresholding

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Image Analysis (AREA)
  • Image Processing (AREA)

Abstract

本发明公开了一种焦平面太赫兹成像中干涉条纹区域识别与处理方法及系统,包括:对焦平面太赫兹成像中得到的图像进行全局阈值分割,获取图像的全局阈值,得到第一图像;判断第一图像是否亮度不均:是则对第一图像进行局部阈值分割得到第二图像,对第二图像进行条纹识别,得到第二图像的暗条纹区域,并将第一图像的暗条纹区域和第二图像的暗条纹区域进行合并并进行条纹消除,得到第三图像;否则对第一图像进行条纹识别,得到第一图像的暗条纹区域并进行条纹消除处理,得到第四图像。对所述第三图像或第四图像进行图像增强处理。本发明能够消除图像中大面积干涉条纹对目标物的干扰的同时还能够大大提高图像的质量。

Figure 202010333623

The invention discloses a method and system for identifying and processing interference fringe areas in focal plane terahertz imaging, comprising: performing global threshold segmentation on an image obtained in focal plane terahertz imaging, acquiring the global threshold of the image, and obtaining a first image; judging Whether the brightness of the first image is uneven: if yes, perform local threshold segmentation on the first image to obtain the second image, perform stripe recognition on the second image, obtain the dark stripe area of the second image, and combine the dark stripe area of the first image and Merge the dark stripe areas of the second image and perform stripe elimination to obtain a third image; otherwise, perform stripe recognition on the first image to obtain the dark stripe areas of the first image and perform stripe elimination processing to obtain a fourth image. performing image enhancement processing on the third image or the fourth image. The invention can eliminate the interference of large-area interference fringes on the target object in the image, and can also greatly improve the image quality.

Figure 202010333623

Description

一种焦平面太赫兹成像中干涉条纹区域识别与处理方法及 系统A method and system for identifying and processing interference fringe areas in focal plane terahertz imaging

技术领域Technical Field

本发明涉及焦平面太赫兹成像领域,特别是涉及一种焦平面太赫兹成像中干涉条纹区域识别与处理方法及系统。The present invention relates to the field of focal plane terahertz imaging, and in particular to a method and system for identifying and processing interference fringe regions in focal plane terahertz imaging.

背景技术Background Art

太赫兹辐射指的是频率在0.1-10THz之间的电磁波,其波段处于微波和红外之间,属于远红外电磁辐射范畴。随着太赫兹波源端和探测端的进步,连续太赫兹成像系统在不断寻求从体积小、速度快、操作方便等方向突破。而太赫兹焦平面阵列探测器因其成本低,操作方便,采集速度快,图像质量信噪比高等优点,应用场景非常广阔。Terahertz radiation refers to electromagnetic waves with a frequency between 0.1-10THz, which is between microwaves and infrared and belongs to the category of far-infrared electromagnetic radiation. With the progress of terahertz wave sources and detectors, continuous terahertz imaging systems are constantly seeking breakthroughs in terms of small size, fast speed, and easy operation. Terahertz focal plane array detectors have a wide range of application scenarios due to their low cost, easy operation, fast acquisition speed, and high image quality signal-to-noise ratio.

但是由于太赫兹波长较长,在从太赫兹波源端到探测端的光路上发生非常明显的干涉现象,导致最后采集的图像背景中出现大面积的干涉条纹,极大地影响目标的成像,且由于太赫兹本身能量较低,图像整体对比度较低,需要对其进行图像增强。在先技术中,一直以来研究人员对此做了大量研究,但大多仍是直接高频滤波或加上带阻滤波器,仅仅只能解决在整张图像上严格规律分布的条纹噪声,一旦干涉条纹仅局部出现或者分布稍稍杂乱一些就无法处理,且常规的图像增强方法同时也会增强条纹的干扰。因此需要一种有效地方法来对太赫兹图像中的条纹进行针对性的识别并处理。However, due to the long wavelength of terahertz, a very obvious interference phenomenon occurs on the optical path from the terahertz wave source to the detection end, resulting in large-area interference fringes in the background of the final collected image, which greatly affects the imaging of the target. In addition, due to the low energy of terahertz itself, the overall contrast of the image is low, and image enhancement is required. In the prior art, researchers have done a lot of research on this, but most of them are still direct high-frequency filtering or adding band-stop filters, which can only solve the stripe noise that is strictly distributed on the entire image. Once the interference fringes appear only locally or the distribution is slightly messy, it cannot be processed, and conventional image enhancement methods will also enhance the interference of the fringes. Therefore, an effective method is needed to identify and process the fringes in terahertz images in a targeted manner.

发明内容Summary of the invention

本发明主要解决的技术问题是提供一种太赫兹图像中的条纹识别方法,能够对太赫兹图像中的条纹进行针对性的识别和处理。The main technical problem solved by the present invention is to provide a method for identifying fringes in a terahertz image, which can perform targeted identification and processing on the fringes in the terahertz image.

为解决上述技术问题,本发明采用的一个技术方案是:提供一种焦平面太赫兹成像中干涉条纹区域识别与处理方法。In order to solve the above technical problems, a technical solution adopted by the present invention is to provide a method for identifying and processing interference fringe areas in focal plane terahertz imaging.

该方法包括以下步骤:The method comprises the following steps:

对焦平面太赫兹成像中得到的图像进行全局阈值分割,获取所述图像的全局阈值,得到第一图像;Performing global threshold segmentation on the image obtained in the focal plane terahertz imaging, obtaining a global threshold of the image, and obtaining a first image;

判断所述第一图像是否亮度不均:是则对所述第一图像进行局部阈值分割得到第二图像,对所述第二图像进行条纹识别,得到所述第二图像的暗条纹区域;否则对所述第一图像进行条纹识别,得到所述第一图像的暗条纹区域;Determine whether the first image has uneven brightness: if yes, perform local threshold segmentation on the first image to obtain a second image, perform stripe recognition on the second image to obtain a dark stripe area of the second image; otherwise, perform stripe recognition on the first image to obtain a dark stripe area of the first image;

若存在亮度不均,将所述第一图像的暗条纹区域和所述第二图像的暗条纹区域进行合并并进行条纹消除处理,得到第三图像;否则对所述第一图像的暗条纹区域进行条纹消除处理,得到第四图像。If there is uneven brightness, the dark stripe area of the first image and the dark stripe area of the second image are merged and stripe elimination processing is performed to obtain a third image; otherwise, the dark stripe area of the first image is subjected to stripe elimination processing to obtain a fourth image.

对所述第三图像或第四图像进行图像增强处理。Perform image enhancement processing on the third image or the fourth image.

优选地,在进行全局阈值分割之前,还包括对所述焦平面太赫兹成像中得到的图像进行线性拉伸。所述全局阈值分割包括使用基于最大类间方差法即OTSU的单阈值分割。Preferably, before performing global threshold segmentation, the image obtained in the focal plane terahertz imaging is linearly stretched. The global threshold segmentation includes using a single threshold segmentation based on the maximum inter-class variance method, namely, OTSU.

优选地,所述判断所述第一图像是否亮度不均包括:Preferably, determining whether the first image has uneven brightness includes:

分别计算所述第一图像上下左右四个边界区域的差异度,所述上下左右四个边界区域指靠近所述第一图像边缘的四个长方形区域;Calculate the difference of four upper, lower, left and right boundary areas of the first image respectively, where the four upper, lower, left and right boundary areas refer to four rectangular areas close to the edge of the first image;

找到所述四个边界区域中最大的差异度;Find the maximum difference among the four boundary regions;

判断所述最大的差异度绝对值若大于阈值时,所述第一图像亮度不均。If it is determined that the maximum absolute value of the difference is greater than a threshold, the brightness of the first image is uneven.

优选地,根据所述亮度不均的情况进行多次局部阈值分割,得到第五图像;对所述第三图像或第四图像进行图像增强处理包括使用增强辅助二值图像,所述增强辅助二值图像包括:Preferably, performing multiple local threshold segmentation according to the uneven brightness to obtain a fifth image; performing image enhancement processing on the third image or the fourth image includes using an enhanced auxiliary binary image, and the enhanced auxiliary binary image includes:

当所述第一图像不存在亮度不均时,所述增强辅助二值图像采用第一图像;When the first image does not have uneven brightness, the enhanced auxiliary binary image uses the first image;

当所述第一图像存在亮度不均,若只进行一次所述局部阈值分割时,所述增强辅助二值图像采用第二图像;否则所述增强辅助二值图像采用所述第五图像。When the first image has uneven brightness, if the local threshold segmentation is performed only once, the enhanced auxiliary binary image uses the second image; otherwise, the enhanced auxiliary binary image uses the fifth image.

优选地,所述最大的差异度为正值时,计算所述第一图像中亮像素的平均坐标位置;所述最大的差异度为负值时,计算所述第一图像中暗像素的平均坐标位置。Preferably, when the maximum difference is a positive value, the average coordinate position of the bright pixels in the first image is calculated; when the maximum difference is a negative value, the average coordinate position of the dark pixels in the first image is calculated.

优选地,所述条纹识别包括:在所述第一图像中,得到特征较好的二值图像,按条纹方向逐行或逐列扫描所述二值图像,保存亮像素点和暗像素点的连续长度至一个长度段中;判断所述长度段中暗条纹的部分,还原成原图像的像素位置并在一个和原图像同样大小的标记矩阵中标记。Preferably, the stripe recognition includes: obtaining a binary image with better features in the first image, scanning the binary image row by row or column by column in the direction of the stripes, and saving the continuous length of bright pixels and dark pixels into a length segment; determining the part of the dark stripes in the length segment, restoring it to the pixel position of the original image and marking it in a marking matrix of the same size as the original image.

优选地,所述条纹识别还包括根据所述标记矩阵中的标记点找到周围一定距离内的相似点。Preferably, the stripe recognition further comprises finding similar points within a certain distance around the marking points in the marking matrix.

优选地,对所述第三图像或第四图像进行图像增强处理包括对所述第三图像或第四图像的背景像素暗条纹区域像素进行灰度提高处理。Preferably, performing image enhancement processing on the third image or the fourth image includes performing grayscale enhancement processing on pixels in a dark stripe area of background pixels of the third image or the fourth image.

优选地,对所述第三图像或第四图像进行所述灰度提高处理后使用均值滤波方式进行降噪处理。Preferably, after the grayscale enhancement process is performed on the third image or the fourth image, a noise reduction process is performed using a mean filtering method.

一种焦平面太赫兹成像中干涉条纹区域识别与处理系统包括:A system for identifying and processing interference fringe regions in focal plane terahertz imaging comprises:

全局阈值分割模块:用于对焦平面太赫兹成像中得到的图像进行全局阈值分割,获取所述图像的全局阈值,得到第一图像;A global threshold segmentation module: used for performing global threshold segmentation on the image obtained in the focal plane terahertz imaging, obtaining a global threshold of the image, and obtaining a first image;

局部阈值分割模块:判断所述第一图像是否亮度不均:是则对所述第一图像进行局部阈值分割得到第二图像,对所述第二图像进行条纹识别,得到所述第二图像的暗条纹区域;否则对所述第一图像进行条纹识别,得到所述第一图像的暗条纹区域;Local threshold segmentation module: determine whether the first image has uneven brightness: if yes, perform local threshold segmentation on the first image to obtain a second image, perform stripe recognition on the second image to obtain a dark stripe area of the second image; otherwise, perform stripe recognition on the first image to obtain a dark stripe area of the first image;

条纹合并和消除模块:若存在亮度不均,将所述第一图像的暗条纹区域和所述第二图像的暗条纹区域进行合并并进行条纹消除处理,得到第三图像;否则对所述第一图像的暗条纹区域进行条纹消除处理,得到第四图像。Stripe merging and elimination module: if there is uneven brightness, the dark stripe area of the first image and the dark stripe area of the second image are merged and stripe elimination is performed to obtain a third image; otherwise, the dark stripe area of the first image is subjected to stripe elimination to obtain a fourth image.

一种焦平面太赫兹成像中干涉条纹区域识别与处理系统还包括:A system for identifying and processing interference fringe regions in focal plane terahertz imaging also includes:

图像增强模块:对所述第三图像或第四图像进行图像增强处理。Image enhancement module: performs image enhancement processing on the third image or the fourth image.

本发明的有益效果是:The beneficial effects of the present invention are:

(1)采用全局阈值分割与局部阈值分割结合的方式,对太赫兹图形进行处理,可以自动识别亮度不均的区域并进行相应的处理;同时准确识别条纹区域并对暗条纹进行处理,消除图像中大面积条纹对目标物的干扰;(2)通过局部阈值分割得到得二值图像进行图像增强,使得目标物更加明显的显现出来;(3)相对于传统的高频滤波加上一些带阻滤波器的处理方法,本方法计算过程更简单,效果更好,且适用面更广,具有巨大的实用价值;(4)一种焦平面太赫兹成像中干涉条纹区域识别与处理系统,能够实时成像且能够最大限度的避免干涉条纹的影响,提高图像的质量。(1) The terahertz image is processed by combining global threshold segmentation with local threshold segmentation, which can automatically identify areas with uneven brightness and perform corresponding processing; at the same time, the stripe area is accurately identified and the dark stripes are processed to eliminate the interference of large-area stripes in the image on the target object; (2) The binary image obtained by local threshold segmentation is enhanced to make the target object more obvious; (3) Compared with the traditional high-frequency filtering plus some band-stop filter processing method, this method has a simpler calculation process, better effect, wider applicability, and great practical value; (4) A system for identifying and processing interference stripe areas in focal plane terahertz imaging can achieve real-time imaging and can avoid the influence of interference stripes to the greatest extent, thereby improving image quality.

附图说明BRIEF DESCRIPTION OF THE DRAWINGS

图1为本发明的总流程图;Fig. 1 is a general flow chart of the present invention;

图2为采集的原始图像;Figure 2 is the original image collected;

图3是经过线性拉升后的图像;Figure 3 is the image after linear stretching;

图4是全局阈值图像分割得到的二值图像;FIG4 is a binary image obtained by global threshold image segmentation;

图5是亮度不均时由一次局部阈值图像分割得到的二值图像;FIG5 is a binary image obtained by segmenting a local threshold image when the brightness is uneven;

图6是用于图像增强辅助的第二次局部阈值分割得到的二值图像;FIG6 is a binary image obtained by the second local threshold segmentation for image enhancement assistance;

图7和图8分别是图4图5经条纹识别结果,其中黑色部分即暗条纹区域;Figures 7 and 8 are the stripe recognition results of Figures 4 and 5, respectively, where the black part is the dark stripe area;

图9是图7和图8的暗条纹区域合并的结果;FIG9 is a result of merging the dark stripe regions of FIG7 and FIG8 ;

图10是暗条纹消除后的效果图;FIG10 is a diagram showing the effect of eliminating dark streaks;

图11是图像增强后的效果图。FIG11 is a diagram showing the effect of image enhancement.

具体实施方式DETAILED DESCRIPTION

为了更好的理解本发明的技术方案,下面将结合说明书附图以及具体的实施方式对本发明技术方案进行详细的说明。In order to better understand the technical solution of the present invention, the technical solution of the present invention will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

实施例1Example 1

在一示例性实施例中,如图1所示一种焦平面太赫兹成像中干涉条纹区域识别与处理方法,包括以下步骤:In an exemplary embodiment, a method for identifying and processing interference fringe regions in focal plane terahertz imaging as shown in FIG1 includes the following steps:

对焦平面太赫兹成像中得到的图像进行全局阈值分割,获取所述图像的全局阈值,得到第一图像;Performing global threshold segmentation on the image obtained in the focal plane terahertz imaging, obtaining a global threshold of the image, and obtaining a first image;

判断所述第一图像是否亮度不均:是则对所述第一图像进行局部阈值分割得到第二图像,对所述第二图像进行条纹识别,得到所述第二图像的暗条纹区域;否则对所述第一图像进行条纹识别,得到所述第一图像的暗条纹区域;Determine whether the first image has uneven brightness: if yes, perform local threshold segmentation on the first image to obtain a second image, perform stripe recognition on the second image to obtain a dark stripe area of the second image; otherwise, perform stripe recognition on the first image to obtain a dark stripe area of the first image;

若存在亮度不均,将所述第一图像的暗条纹区域和所述第二图像的暗条纹区域进行合并并进行条纹消除处理,得到第三图像;否则对所述第一图像的暗条纹区域进行条纹消除处理,得到第四图像。If there is uneven brightness, the dark stripe area of the first image and the dark stripe area of the second image are merged and stripe elimination processing is performed to obtain a third image; otherwise, the dark stripe area of the first image is subjected to stripe elimination processing to obtain a fourth image.

对所述第三图像或第四图像进行图像增强处理。Perform image enhancement processing on the third image or the fourth image.

现有技术中,.焦平面太赫兹图像的分割难点是对比度差,图像整体模糊不容易看清楚形状且有条纹的影响,同时光照分布不均匀,本发明结合全局阈值分割与局部阈值分割,具体结合焦平面太赫兹图像的特点,基于干涉条纹的空间分布规律直接识别干涉条纹在图像中的对应位置从而针对性消除。In the prior art, the difficulty in segmenting focal plane terahertz images is poor contrast. The overall image is blurred and it is not easy to see the shape clearly. There is also the influence of stripes. At the same time, the illumination distribution is uneven. The present invention combines global threshold segmentation with local threshold segmentation, specifically combines the characteristics of focal plane terahertz images, and directly identifies the corresponding positions of interference fringes in the image based on the spatial distribution law of interference fringes, so as to eliminate them in a targeted manner.

优选地,在进行全局阈值分割之前,还包括对所述焦平面太赫兹成像中得到的图像进行线性拉伸。所述全局阈值分割包括使用基于最大类间方差法即OTSU的单阈值分割。Preferably, before performing global threshold segmentation, the image obtained in the focal plane terahertz imaging is linearly stretched. The global threshold segmentation includes using a single threshold segmentation based on the maximum inter-class variance method, namely, OTSU.

优选地,所述判断所述第一图像是否亮度不均包括:Preferably, determining whether the first image has uneven brightness includes:

分别计算所述第一图像上下左右四个边界区域的差异度,所述上下左右四个边界区域指靠近所述第一图像边缘的四个长方形区域;Calculate the difference of four upper, lower, left and right boundary areas of the first image respectively, where the four upper, lower, left and right boundary areas refer to four rectangular areas close to the edge of the first image;

找到所述四个边界区域中最大的差异度;Find the maximum difference among the four boundary regions;

判断所述最大的差异度绝对值若大于阈值时,所述第一图像亮度不均。If it is determined that the maximum absolute value of the difference is greater than a threshold, the brightness of the first image is uneven.

进一步地,分别计算所述图像上下左右四个边界区域在全局阈值分别加减20的情况下的差异度。大小为M*N的图像中,四个边界区域分别为,0<x<M/10,0<y<N/10,0.9×M<x<M,0.9×N<y<N,差异度P的计算方法为:Furthermore, the difference between the four boundary areas of the image is calculated respectively when the global threshold is plus or minus 20. In an image of size M*N, the four boundary areas are 0<x<M/10, 0<y<N/10, 0.9×M<x<M, 0.9×N<y<N, and the difference P is calculated as follows:

Figure BDA0002465828120000061
Figure BDA0002465828120000061

其中,l1为在全局阈值加20的情况下区域二值化后亮像素的个数,d2为在全局阈值减20的情况下区域二值化后暗像素的个数,l和d是阈值没有改变的情况下亮像素和暗像素的个数。Among them, l1 is the number of bright pixels after regional binarization when the global threshold is plus 20, d2 is the number of dark pixels after regional binarization when the global threshold is minus 20, and l and d are the numbers of bright pixels and dark pixels when the threshold is not changed.

进一步地,所述第一图像是否亮度不均还包括找到所述四个边界区域中最大的差异度,所述最大的差异度绝对值大于0.5时,所述第一图像亮度不均。Furthermore, determining whether the first image has uneven brightness further includes finding the maximum difference among the four boundary areas, and when the absolute value of the maximum difference is greater than 0.5, the first image has uneven brightness.

进一步地,根据所述亮度不均的情况进行多次局部阈值分割。所述S2中进行局部阈值分割时对所述S1中全局阈值进行阈值调整。Furthermore, multiple local threshold segmentations are performed according to the uneven brightness. When the local threshold segmentation is performed in S2, the global threshold in S1 is adjusted.

进一步地,所述最大的差异度为正值时,计算所述第一图像中亮像素的平均坐标位置p0;所述最大的差异度为负值时,计算所述第一图像中暗像素的平均坐标位置。记暗像素坐标位置为D(a,b),将D到p0的原取值区域U垂直方向上的阈值进行调整。Furthermore, when the maximum difference is a positive value, the average coordinate position p 0 of the bright pixels in the first image is calculated; when the maximum difference is a negative value, the average coordinate position of the dark pixels in the first image is calculated. The dark pixel coordinate position is recorded as D(a,b), and the threshold in the vertical direction of the original value range U from D to p 0 is adjusted.

当这个方向水平时,调整的公式如下:When this direction is horizontal, the adjustment formula is as follows:

Figure BDA0002465828120000062
min(l,a)<i≤max(l,a)
Figure BDA0002465828120000062
min(l, a)<i≤max(l, a)

当这个方向竖直时,调整的公式如下:When this direction is vertical, the adjustment formula is as follows:

Figure BDA0002465828120000063
min(l,b)<j≤max(l,b)
Figure BDA0002465828120000063
min(l, b)<j≤max(l, b)

其中,T为调整后的阈值,T0为全局阈值,Q为调整的幅度,当水平向左或竖直向上时L=0,水平向右时L=M,竖直向下时L=N,M,N为图像长宽。Wherein, T is the adjusted threshold, T0 is the global threshold, Q is the adjustment amplitude, L=0 when horizontally to the left or vertically upward, L=M when horizontally to the right, and L=N when vertically downward, and M and N are the length and width of the image.

进一步地,根据所述亮度不均的情况进行多次局部阈值分割,得到第五图像;对所述第三图像或第四图像进行图像增强处理包括使用增强辅助二值图像,所述增强辅助二值图像包括:Further, performing multiple local threshold segmentation according to the uneven brightness to obtain a fifth image; performing image enhancement processing on the third image or the fourth image includes using an enhanced auxiliary binary image, and the enhanced auxiliary binary image includes:

当所述第一图像不存在亮度不均时,所述增强辅助二值图像采用第一图像;When the first image does not have uneven brightness, the enhanced auxiliary binary image uses the first image;

当所述第一图像存在亮度不均,若只进行一次所述局部阈值分割时,所述增强辅助二值图像采用第二图像;否则所述增强辅助二值图像采用所述第五图像。When the first image has uneven brightness, if the local threshold segmentation is performed only once, the enhanced auxiliary binary image uses the second image; otherwise, the enhanced auxiliary binary image uses the fifth image.

进一步地,若U的反向区域的差异度p1的绝对值大于0.3时,还需进行一次从D到p1的局部阈值调整过程。得到的二值图像用于图像增强过程中的增强辅助二值图像。Furthermore, if the absolute value of the difference degree p 1 of the reverse region of U is greater than 0.3, a local threshold adjustment process from D to p 1 is required. The obtained binary image is used as an enhanced auxiliary binary image in the image enhancement process.

进一步地,所述条纹识别包括:在所述第一图像中,得到特征较好的二值图像,按条纹方向逐行或逐列扫描所述二值图像,保存亮像素点和暗像素点的连续长度至一个长度段中;判断所述长度段中暗条纹的部分,还原成原图像的像素位置并在一个和原图像同样大小的标记矩阵中标记。Furthermore, the stripe recognition includes: obtaining a binary image with better features in the first image, scanning the binary image row by row or column by column in the direction of the stripes, and saving the continuous length of bright pixels and dark pixels into a length segment; determining the part of the dark stripes in the length segment, restoring it to the pixel position of the original image and marking it in a marking matrix of the same size as the original image.

进一步地,所述条纹识别还包括根据所述标记矩阵中的标记点找到周围一定距离内的相似点。Furthermore, the stripe recognition also includes finding similar points within a certain distance around the marked points in the marking matrix.

进一步地,所述第三图像进行图像增强处理包括对所述第三图像的背景像素暗条纹区域像素进行灰度提高处理。Furthermore, the image enhancement processing of the third image includes performing grayscale enhancement processing on the background pixels and dark stripe area pixels of the third image.

进一步地,对所述第三图像进行所述灰度提高处理后使用均值滤波方式进行降噪处理。Furthermore, after the grayscale enhancement process is performed on the third image, a noise reduction process is performed using a mean filtering method.

实施例2Example 2

一种焦平面太赫兹成像中干涉条纹区域识别与处理系统包括:A system for identifying and processing interference fringe regions in focal plane terahertz imaging comprises:

全局阈值分割模块:用于对焦平面太赫兹成像中得到的图像进行全局阈值分割,获取所述图像的全局阈值,得到第一图像;A global threshold segmentation module: used for performing global threshold segmentation on the image obtained in the focal plane terahertz imaging, obtaining a global threshold of the image, and obtaining a first image;

局部阈值分割模块:判断所述第一图像是否亮度不均:是则对所述第一图像进行局部阈值分割得到第二图像,对所述第二图像进行条纹识别,得到所述第二图像的暗条纹区域;否则对所述第一图像进行条纹识别,得到所述第一图像的暗条纹区域;Local threshold segmentation module: determine whether the first image has uneven brightness: if yes, perform local threshold segmentation on the first image to obtain a second image, perform stripe recognition on the second image to obtain a dark stripe area of the second image; otherwise, perform stripe recognition on the first image to obtain a dark stripe area of the first image;

条纹合并和消除模块:若存在亮度不均,将所述第一图像的暗条纹区域和所述第二图像的暗条纹区域进行合并并进行条纹消除处理,得到第三图像;否则对所述第一图像的暗条纹区域进行条纹消除处理,得到第四图像。Stripe merging and elimination module: if there is uneven brightness, the dark stripe area of the first image and the dark stripe area of the second image are merged and stripe elimination is performed to obtain a third image; otherwise, the dark stripe area of the first image is subjected to stripe elimination to obtain a fourth image.

图像增强模块:对所述第三图像或第四图像进行图像增强处理。Image enhancement module: performs image enhancement processing on the third image or the fourth image.

具体地,在一示例性实施例中,在使用较小焦平面太赫兹探测器扫描成像系统对刀片成像后,得到图2所示的原图像,可以看出这个图像右半部分有着规律的明暗相间条纹,这个是由太赫兹波源端到探测端的光路中发生干涉引起的,同时图像对比度低,难以识别图像中的刀片。本发明就此提出一种焦平面太赫兹成像中干涉条纹区域识别与处理方法,流程图如图1所示,包括如下过程:Specifically, in an exemplary embodiment, after imaging the blade using a smaller focal plane terahertz detector scanning imaging system, the original image shown in FIG2 is obtained. It can be seen that the right half of the image has regular light and dark stripes, which is caused by interference in the optical path from the terahertz wave source end to the detection end. At the same time, the image contrast is low, and it is difficult to identify the blade in the image. The present invention proposes a method for identifying and processing interference stripe areas in focal plane terahertz imaging. The flow chart is shown in FIG1, and includes the following processes:

线性拉升过程:对图2使用直方图线性拉升提高对比度,可以一定程度上提高阈值分割的准确性,结果如图3所示。Linear pull-up process: Using the histogram linear pull-up to improve the contrast of Figure 2 can improve the accuracy of threshold segmentation to a certain extent. The result is shown in Figure 3.

全局阈值分割过程:对图3使用基于最大类间方差法(OTSU)的单阈值图像分割来获取能够较好地体现条纹与物体特征的二值图像,如图4所示,黑色部分理想状态下应该包括实际成像目标及暗条纹,白色部分则为背景。同时还需要保留全局阈值,用于之后的局部阈值分割和条纹消除。Global threshold segmentation process: Single threshold image segmentation based on the maximum between-class variance method (OTSU) is used for Figure 3 to obtain a binary image that can better reflect the characteristics of stripes and objects, as shown in Figure 4. Ideally, the black part should include the actual imaging target and dark stripes, and the white part is the background. At the same time, the global threshold needs to be retained for subsequent local threshold segmentation and stripe elimination.

局部阈值分割过程:根据是否发生亮度不均来判断是否使用一种新的局部阈值分割方法,来得到改善的二值图像,如图5所示,新的二值图像能够将图像下半部分的目标与背景分割出来,更好的展现条纹与刀片同背景的区别;对于图像增强所需的二次使用局部阈值分割的结果,就如图6所示,可以清晰地看到整个刀片实际应有的形状了。Local threshold segmentation process: Depending on whether uneven brightness occurs, it is determined whether to use a new local threshold segmentation method to obtain an improved binary image, as shown in Figure 5. The new binary image can separate the target from the background in the lower half of the image, and better show the difference between the stripes and the blade and the background. The result of the secondary use of local threshold segmentation required for image enhancement is shown in Figure 6, and the actual shape of the entire blade can be clearly seen.

而亮度不均的判断方法具体为,分别计算整个图像上下左右四个边界区域在全局阈值分别加减20的情况下的差异度,大小为M*N的图像中,四个边界区域分别为,0<x<M/10,0<y<N/10,0.9×M<x<M,0.9×N<y<N,差异度P的计算方法为:The specific method for judging uneven brightness is to calculate the difference between the four boundary areas of the whole image when the global threshold is added or subtracted by 20. In an image of size M*N, the four boundary areas are 0<x<M/10, 0<y<N/10, 0.9×M<x<M, 0.9×N<y<N, and the calculation method of the difference P is:

Figure BDA0002465828120000091
Figure BDA0002465828120000091

其中,l1为在全局阈值加20的情况下区域二值化后亮像素的个数,d2为在全局阈值减20的情况下区域二值化后暗像素的个数,l和d是阈值没有改变的情况下亮像素和暗像素的个数。找到四个区域中最大的差异度,当最大的差异度绝对值大于0.5的时候即可判断图像中有亮度不均的现象。Among them, l1 is the number of bright pixels after regional binarization when the global threshold is plus 20, d2 is the number of dark pixels after regional binarization when the global threshold is minus 20, l and d are the number of bright pixels and dark pixels when the threshold is not changed. Find the largest difference among the four regions. When the absolute value of the largest difference is greater than 0.5, it can be judged that there is uneven brightness in the image.

局部阈值分割方法具体为,在全局阈值分割得到的图4中,当最大的差异度为正值时计算亮像素的平均坐标位置p0,否则计算暗像素的平均坐标位置,记这个坐标位置为D(a,b),将D到p0的原取值区域U垂直方向上的阈值进行调整。The local threshold segmentation method is specifically as follows: in Figure 4 obtained by global threshold segmentation, when the maximum difference is a positive value, the average coordinate position p 0 of the bright pixels is calculated; otherwise, the average coordinate position of the dark pixels is calculated and recorded as D(a,b). The threshold in the vertical direction of the original value area U from D to p 0 is adjusted.

当这个方向水平时,调整的公式如下:When this direction is horizontal, the adjustment formula is as follows:

Figure BDA0002465828120000092
min(l,a)<i≤max(l,a)
Figure BDA0002465828120000092
min(l, a)<i≤max(l, a)

当这个方向竖直时,调整的公式如下:When this direction is vertical, the adjustment formula is as follows:

Figure BDA0002465828120000093
min(l,b)<j≤max(l,b)
Figure BDA0002465828120000093
min(l, b)<j≤max(l, b)

其中,T为调整后的阈值,T0为全局阈值,Q为调整的幅度,当水平向左或竖直向上时L=0,水平向右时L=M,竖直向下时L=N,M,N为图像长宽。Wherein, T is the adjusted threshold, T0 is the global threshold, Q is the adjustment amplitude, L=0 when horizontally to the left or vertically upward, L=M when horizontally to the right, and L=N when vertically downward, and M and N are the length and width of the image.

特别的,为得到图像增强步骤里所需的二值图像,若U的反向区域的差异度p1的绝对值大于0.3时,还需进行一次从D到p1的局部阈值调整过程。In particular, in order to obtain the binary image required in the image enhancement step, if the absolute value of the difference p 1 in the reverse region of U is greater than 0.3, a local threshold adjustment process from D to p 1 is required.

条纹识别过程:在二值图像中根据光的干涉规律,从明暗条纹的宽度,方向和级数来确定干涉区域,识别暗条纹部分。从图3可以看出,刀片右边的条纹明显符合干涉条纹的规律。图7和图8则分别是图5图6经条纹识别后的结果,黑色部分既为原图中识别到的干涉暗条纹部分。识别方法具体为:Stripe recognition process: In the binary image, according to the interference law of light, the interference area is determined from the width, direction and level of the light and dark stripes, and the dark stripe part is identified. As can be seen from Figure 3, the stripes on the right side of the blade clearly conform to the law of interference stripes. Figures 7 and 8 are the results of Figures 5 and 6 after stripe recognition, and the black part is the interference dark stripe part identified in the original image. The specific recognition method is:

第一步,在第一图像中得到特征较好的二值图像,按条纹方向逐行或逐列扫描二值图像,保存亮像素点和暗像素点的连续长度至一个长度段中。In the first step, a binary image with better features is obtained in the first image, and the binary image is scanned row by row or column by column in the stripe direction, and the continuous lengths of bright pixels and dark pixels are saved in a length segment.

第二步,依据光的干涉条纹中明暗条纹宽度间距大致相同,且条纹级数一般较多的空间分布规律,判断长度段中暗条纹的部分,还原成原图像的像素位置并在一个和原图像同样大小的标记矩阵中标记。In the second step, based on the spatial distribution law that the width and spacing of light and dark stripes in the interference fringes of light are roughly the same and the number of stripes is generally large, the part of the dark stripes in the length segment is determined, restored to the pixel position of the original image and marked in a marking matrix of the same size as the original image.

第三步,根据标记矩阵中的标记点找到周围一定距离内的相似点,这些相似点具有和标记点灰度值大小相似且连续的特性,这样就能识别出暗条纹的大部分区域。The third step is to find similar points within a certain distance around the marked points in the marking matrix. These similar points have the characteristics of similar and continuous grayscale values as the marked points, so that most areas of the dark stripes can be identified.

区域合并过程:合并由全局阈值分割和局部阈值分割的二值图像得到的暗条纹区域,合并的条纹区域如图9所示,基本覆盖了实际暗条纹区域。Region merging process: The dark stripe region obtained by merging the binary images segmented by global threshold segmentation and local threshold segmentation is shown in FIG9 , which basically covers the actual dark stripe region.

条纹消除过程:对上述合并暗条纹区域,通过全局阈值和局部阈值调整条纹区域的灰度值,从而消除条纹的影响。调整条纹区域的灰度值的公式为:G=T1+T2-G0,其中,T1是图像的全局阈值,T2是图像的局部阈值,G0是原图像灰度值;G是变换后的灰度值。最终的结果如图10所示,可以看出干涉条纹对图像的影响基本消除,但由于太赫兹图像对比度低的问题,还需要对图像进一步处理。Stripe elimination process: For the above-mentioned merged dark stripe area, the grayscale value of the stripe area is adjusted by the global threshold and the local threshold to eliminate the influence of the stripe. The formula for adjusting the grayscale value of the stripe area is: G = T 1 + T 2 - G 0 , where T1 is the global threshold of the image, T2 is the local threshold of the image, G 0 is the grayscale value of the original image; G is the grayscale value after transformation. The final result is shown in Figure 10. It can be seen that the influence of interference fringes on the image is basically eliminated, but due to the low contrast of the terahertz image, the image needs to be further processed.

图像增强过程:由于焦平面太赫兹图像的质量较差且噪声严重且复杂,传统的线性拉升与直方图均衡化都不能达到理想的效果,这里我们创新性的结合局部阈值分割结果来提高图像对比度。将第三图像的背景像素暗条纹区域像素进行灰度提高处理,其余部分的灰度降低,最后通过均值滤波降低噪声。最终结果如图11所示,极大消除了条纹影响的同时可以在图像中比较清晰得看出刀片的实际形状,且给人的直观视觉效果大大提高,而这个效果是所有滤波方法不可能得到的。Image enhancement process: Due to the poor quality of focal plane terahertz images and the serious and complex noise, traditional linear pull-up and histogram equalization cannot achieve the desired effect. Here we innovatively combine the local threshold segmentation results to improve the image contrast. The grayscale of the dark stripe area of the background pixels of the third image is increased, and the grayscale of the rest is reduced. Finally, the noise is reduced by mean filtering. The final result is shown in Figure 11. While greatly eliminating the influence of stripes, the actual shape of the blade can be seen more clearly in the image, and the intuitive visual effect is greatly improved, which is impossible to obtain by all filtering methods.

以上所述仅为本发明的实施例,并非因此限制本发明的专利范围,凡是利用本发明说明书及附图内容所作的等效结构或等效流程变换,或直接或间接运用在其他相关的技术领域,均同理包括在本发明的专利保护范围内。The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims (9)

1.一种焦平面太赫兹成像中干涉条纹区域识别与处理方法,其特征在于:包括以下步骤:1. A method for identifying and processing interference fringe areas in focal plane terahertz imaging, characterized in that it comprises the following steps: 对焦平面太赫兹成像中得到的图像进行全局阈值分割,获取所述图像的全局阈值,得到第一图像;Performing global threshold segmentation on the image obtained in the focal plane terahertz imaging, obtaining a global threshold of the image, and obtaining a first image; 判断所述第一图像是否亮度不均:是则对所述第一图像进行局部阈值分割得到第二图像,对所述第二图像进行条纹识别,得到所述第二图像的暗条纹区域;否则对所述第一图像进行条纹识别,得到所述第一图像的暗条纹区域;所述判断所述第一图像是否亮度不均包括:Determine whether the first image has uneven brightness: if yes, perform local threshold segmentation on the first image to obtain a second image, perform stripe recognition on the second image to obtain a dark stripe area of the second image; otherwise, perform stripe recognition on the first image to obtain a dark stripe area of the first image; the determining whether the first image has uneven brightness includes: 分别计算所述第一图像上下左右四个边界区域的差异度,所述上下左右四个边界区域指靠近所述第一图像边缘的四个长方形区域;Calculate the difference of four upper, lower, left and right boundary areas of the first image respectively, where the four upper, lower, left and right boundary areas refer to four rectangular areas close to the edge of the first image; 找到所述四个边界区域中最大的差异度;Find the maximum difference among the four boundary regions; 判断所述最大的差异度绝对值若大于阈值时,所述第一图像亮度不均;If it is determined that the maximum absolute value of the difference is greater than a threshold, the brightness of the first image is uneven; 根据所述亮度不均的情况进行多次局部阈值分割;在进行局部阈值分割时对全局阈值进行阈值调整,当所述最大的差异度为正值时,计算所述第一图像中亮像素的平均坐标位置p0;所述最大的差异度为负值时,计算所述第一图像中暗像素的平均坐标位置;Performing multiple local threshold segmentations according to the uneven brightness; adjusting the global threshold when performing local threshold segmentation, and calculating the average coordinate position p 0 of bright pixels in the first image when the maximum difference is a positive value; and calculating the average coordinate position of dark pixels in the first image when the maximum difference is a negative value; 记暗像素坐标位置为D(a,b),将D到p0的原取值区域U垂直方向上的阈值进行调整;The dark pixel coordinate position is recorded as D(a,b), and the threshold in the vertical direction of the original value range U from D to p 0 is adjusted; 当这个方向水平时,调整的公式如下:When this direction is horizontal, the adjustment formula is as follows:
Figure QLYQS_1
Figure QLYQS_1
当这个方向竖直时,调整的公式如下:When this direction is vertical, the adjustment formula is as follows:
Figure QLYQS_2
Figure QLYQS_2
其中,T为调整后的阈值,T0为全局阈值,Q为调整的幅度,当水平向左或竖直向上时L=0,水平向右时L=M,竖直向下时L=N,M,N为图像长宽;Where T is the adjusted threshold, T 0 is the global threshold, Q is the adjustment amplitude, L=0 when horizontally to the left or vertically upward, L=M when horizontally to the right, L=N when vertically downward, M, N are the length and width of the image; 若存在亮度不均,将所述第一图像的暗条纹区域和所述第二图像的暗条纹区域进行合并并进行条纹消除处理,得到第三图像;否则对所述第一图像的暗条纹区域进行条纹消除处理,得到第四图像;If there is uneven brightness, the dark stripe region of the first image and the dark stripe region of the second image are combined and stripe elimination is performed to obtain a third image; otherwise, the dark stripe region of the first image is subjected to stripe elimination to obtain a fourth image; 对所述第三图像或第四图像进行图像增强处理。Perform image enhancement processing on the third image or the fourth image.
2.根据权利要求1所述的一种焦平面太赫兹成像中干涉条纹区域识别与处理方法,其特征在于:在进行全局阈值分割之前,还包括对所述焦平面太赫兹成像中得到的图像进行线性拉伸。2. The method for identifying and processing interference fringe areas in focal plane terahertz imaging according to claim 1, characterized in that: before performing global threshold segmentation, it also includes linear stretching of the image obtained in the focal plane terahertz imaging. 3.根据权利要求1所述的一种焦平面太赫兹成像中干涉条纹区域识别与处理方法,其特征在于:所述全局阈值分割包括使用基于最大类间方差法即OTSU的单阈值分割。3. The method for identifying and processing interference fringe areas in focal plane terahertz imaging according to claim 1, wherein the global threshold segmentation includes using a single threshold segmentation based on the maximum between-class variance method (OTSU). 4.根据权利要求1所述的一种焦平面太赫兹成像中干涉条纹区域识别与处理方法,其特征在于:根据所述亮度不均的情况进行多次局部阈值分割,得到第五图像;对所述第三图像或第四图像进行图像增强处理包括使用增强辅助二值图像,所述增强辅助二值图像包括:4. The method for identifying and processing interference fringe areas in focal plane terahertz imaging according to claim 1, characterized in that: performing multiple local threshold segmentations according to the uneven brightness to obtain a fifth image; performing image enhancement processing on the third image or the fourth image includes using an enhanced auxiliary binary image, and the enhanced auxiliary binary image includes: 当所述第一图像不存在亮度不均时,所述增强辅助二值图像采用第一图像;When the first image does not have uneven brightness, the enhanced auxiliary binary image uses the first image; 当所述第一图像存在亮度不均,若只进行一次所述局部阈值分割时,所述增强辅助二值图像采用第二图像;否则所述增强辅助二值图像采用所述第五图像。When the first image has uneven brightness, if the local threshold segmentation is performed only once, the enhanced auxiliary binary image uses the second image; otherwise, the enhanced auxiliary binary image uses the fifth image. 5.根据权利要求1所述的一种焦平面太赫兹成像中干涉条纹区域识别与处理方法,其特征在于:所述最大的差异度为正值时,计算所述第一图像中亮像素的平均坐标位置;所述最大的差异度为负值时,计算所述第一图像中暗像素的平均坐标位置。5. The method for identifying and processing interference fringe areas in focal plane terahertz imaging according to claim 1, wherein when the maximum difference is a positive value, the average coordinate position of bright pixels in the first image is calculated; when the maximum difference is a negative value, the average coordinate position of dark pixels in the first image is calculated. 6.根据权利要求1所述的一种焦平面太赫兹成像中干涉条纹区域识别与处理方法,其特征在于:对所述第三图像或第四图像进行图像增强处理包括对所述第三图像或第四图像的背景像素暗条纹区域像素进行灰度提高处理。6. A method for identifying and processing interference fringe areas in focal plane terahertz imaging according to claim 1, characterized in that: performing image enhancement processing on the third image or the fourth image includes performing grayscale enhancement processing on pixels in dark fringe areas of background pixels of the third image or the fourth image. 7.根据权利要求6所述的一种焦平面太赫兹成像中干涉条纹区域识别与处理方法,其特征在于:对所述第三图像或第四图像进行所述灰度提高处理后使用均值滤波方式进行降噪处理。7. The method for identifying and processing interference fringe areas in focal plane terahertz imaging according to claim 6, characterized in that: after the grayscale enhancement processing is performed on the third image or the fourth image, a mean filtering method is used to perform noise reduction processing. 8.一种焦平面太赫兹成像中干涉条纹区域识别与处理系统,其特征在于:系统包括:8. A system for identifying and processing interference fringe areas in focal plane terahertz imaging, characterized in that the system comprises: 全局阈值分割模块:用于对焦平面太赫兹成像中得到的图像进行全局阈值分割,获取所述图像的全局阈值,得到第一图像;A global threshold segmentation module: used for performing global threshold segmentation on the image obtained in the focal plane terahertz imaging, obtaining a global threshold of the image, and obtaining a first image; 局部阈值分割模块:判断所述第一图像是否亮度不均:是则对所述第一图像进行局部阈值分割得到第二图像,对所述第二图像进行条纹识别,得到所述第二图像的暗条纹区域;否则对所述第一图像进行条纹识别,得到所述第一图像的暗条纹区域;所述判断所述第一图像是否亮度不均包括:Local threshold segmentation module: determine whether the first image has uneven brightness: if yes, perform local threshold segmentation on the first image to obtain a second image, perform stripe recognition on the second image to obtain a dark stripe area of the second image; otherwise, perform stripe recognition on the first image to obtain a dark stripe area of the first image; the determination of whether the first image has uneven brightness includes: 分别计算所述第一图像上下左右四个边界区域的差异度,所述上下左右四个边界区域指靠近所述第一图像边缘的四个长方形区域;Calculate the difference of four upper, lower, left and right boundary areas of the first image respectively, where the four upper, lower, left and right boundary areas refer to four rectangular areas close to the edge of the first image; 找到所述四个边界区域中最大的差异度;Find the maximum difference among the four boundary regions; 判断所述最大的差异度绝对值若大于阈值时,所述第一图像亮度不均;If it is determined that the maximum absolute value of the difference is greater than a threshold, the brightness of the first image is uneven; 根据所述亮度不均的情况进行多次局部阈值分割;在进行局部阈值分割时对全局阈值进行阈值调整,当所述最大的差异度为正值时,计算所述第一图像中亮像素的平均坐标位置p0;所述最大的差异度为负值时,计算所述第一图像中暗像素的平均坐标位置;Performing multiple local threshold segmentations according to the uneven brightness; adjusting the global threshold when performing local threshold segmentation, and calculating the average coordinate position p 0 of bright pixels in the first image when the maximum difference is a positive value; and calculating the average coordinate position of dark pixels in the first image when the maximum difference is a negative value; 记暗像素坐标位置为D(a,b),将D到p0的原取值区域U垂直方向上的阈值进行调整;The dark pixel coordinate position is recorded as D(a,b), and the threshold in the vertical direction of the original value range U from D to p 0 is adjusted; 当这个方向水平时,调整的公式如下:When this direction is horizontal, the adjustment formula is as follows:
Figure QLYQS_3
Figure QLYQS_3
当这个方向竖直时,调整的公式如下:When this direction is vertical, the adjustment formula is as follows:
Figure QLYQS_4
Figure QLYQS_4
其中,T为调整后的阈值,T0为全局阈值,Q为调整的幅度,当水平向左或竖直向上时L=0,水平向右时L=M,竖直向下时L=N,M,N为图像长宽;Where T is the adjusted threshold, T 0 is the global threshold, Q is the adjustment amplitude, L=0 when horizontally to the left or vertically upward, L=M when horizontally to the right, L=N when vertically downward, M, N are the length and width of the image; 条纹合并和消除模块:若存在亮度不均,将所述第一图像的暗条纹区域和所述第二图像的暗条纹区域进行合并并进行条纹消除处理,得到第三图像;否则对所述第一图像的暗条纹区域进行条纹消除处理,得到第四图像。Stripe merging and elimination module: if there is uneven brightness, the dark stripe area of the first image and the dark stripe area of the second image are merged and stripe elimination is performed to obtain a third image; otherwise, the dark stripe area of the first image is subjected to stripe elimination to obtain a fourth image.
9.根据权利要求8所述的一种焦平面太赫兹成像中干涉条纹区域识别与处理系统,其特征在于:系统还包括:9. The interference fringe region recognition and processing system in focal plane terahertz imaging according to claim 8, characterized in that the system further comprises: 图像增强模块:对所述第三图像或第四图像进行图像增强处理。Image enhancement module: performs image enhancement processing on the third image or the fourth image.
CN202010333623.6A 2020-04-24 2020-04-24 Method and system for identifying and processing interference fringe region in terahertz imaging of focal plane Active CN111539967B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010333623.6A CN111539967B (en) 2020-04-24 2020-04-24 Method and system for identifying and processing interference fringe region in terahertz imaging of focal plane

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010333623.6A CN111539967B (en) 2020-04-24 2020-04-24 Method and system for identifying and processing interference fringe region in terahertz imaging of focal plane

Publications (2)

Publication Number Publication Date
CN111539967A CN111539967A (en) 2020-08-14
CN111539967B true CN111539967B (en) 2023-03-28

Family

ID=71977256

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010333623.6A Active CN111539967B (en) 2020-04-24 2020-04-24 Method and system for identifying and processing interference fringe region in terahertz imaging of focal plane

Country Status (1)

Country Link
CN (1) CN111539967B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112179505B (en) * 2020-09-23 2022-08-02 中国科学院光电技术研究所 A device and method for image processing based on wedge-shaped plate shear interferometer
CN112489058B (en) * 2020-12-03 2024-08-23 山东承势电子科技有限公司 Efficient and accurate stripe direction estimation method
CN119063659B (en) * 2024-11-05 2025-01-24 成都中住光纤有限公司 Device and method for identifying concentricity of optical fiber coating

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103940510A (en) * 2014-04-02 2014-07-23 中国科学院紫金山天文台 Device and method for detecting terahertz signal multi-dimensional image through dual fourier transformation
CN104517269A (en) * 2014-12-25 2015-04-15 深圳市一体太赫兹科技有限公司 Terahertz image strip processing method and system
CN104574300A (en) * 2014-12-25 2015-04-29 深圳市一体太赫兹科技有限公司 Terahertz image noise processing method and system
CN107095673A (en) * 2017-05-24 2017-08-29 西北核技术研究所 A kind of reflective real-time tumor imaging method and system
CN107481205A (en) * 2017-08-23 2017-12-15 电子科技大学 A kind of Terahertz image fringes noise processing method and system
CN108537735A (en) * 2018-04-16 2018-09-14 电子科技大学 A kind of image split-joint method of focal plane terahertz imaging
CN108956645A (en) * 2018-07-18 2018-12-07 丹阳市精通眼镜技术创新服务中心有限公司 A kind of the optical mirror slip defect detecting device and method of more vision systems
CN109029727A (en) * 2018-06-11 2018-12-18 中国科学院紫金山天文台 Highly sensitive Terahertz superconduction spectrum imaging system and system imaging method based on code aperture
CN110196462A (en) * 2019-06-28 2019-09-03 苏州晶萃光学科技有限公司 A kind of Terahertz lens, preparation method and the terahertz optics system comprising it
CN110443807A (en) * 2019-06-27 2019-11-12 中国地质大学(武汉) A kind of even carrying out image threshold segmentation method of uneven illumination based on luminance proportion
US10568519B1 (en) * 2014-06-02 2020-02-25 The Spectranetics Corporation Terahertz scanning system for an intravascular space

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100059677A1 (en) * 2008-09-10 2010-03-11 Rainer Leonhardt Aspheric Lenses for Imaging

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103940510A (en) * 2014-04-02 2014-07-23 中国科学院紫金山天文台 Device and method for detecting terahertz signal multi-dimensional image through dual fourier transformation
US10568519B1 (en) * 2014-06-02 2020-02-25 The Spectranetics Corporation Terahertz scanning system for an intravascular space
CN104517269A (en) * 2014-12-25 2015-04-15 深圳市一体太赫兹科技有限公司 Terahertz image strip processing method and system
CN104574300A (en) * 2014-12-25 2015-04-29 深圳市一体太赫兹科技有限公司 Terahertz image noise processing method and system
CN107095673A (en) * 2017-05-24 2017-08-29 西北核技术研究所 A kind of reflective real-time tumor imaging method and system
CN107481205A (en) * 2017-08-23 2017-12-15 电子科技大学 A kind of Terahertz image fringes noise processing method and system
CN108537735A (en) * 2018-04-16 2018-09-14 电子科技大学 A kind of image split-joint method of focal plane terahertz imaging
CN109029727A (en) * 2018-06-11 2018-12-18 中国科学院紫金山天文台 Highly sensitive Terahertz superconduction spectrum imaging system and system imaging method based on code aperture
CN108956645A (en) * 2018-07-18 2018-12-07 丹阳市精通眼镜技术创新服务中心有限公司 A kind of the optical mirror slip defect detecting device and method of more vision systems
CN110443807A (en) * 2019-06-27 2019-11-12 中国地质大学(武汉) A kind of even carrying out image threshold segmentation method of uneven illumination based on luminance proportion
CN110196462A (en) * 2019-06-28 2019-09-03 苏州晶萃光学科技有限公司 A kind of Terahertz lens, preparation method and the terahertz optics system comprising it

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Terahertz Imaging Based on Morphological Reconstruction;Jia Shi等;《 IEEE Journal of Selected Topics in Quantum Electronics 》;20170109;第1-7页 *
基于太赫兹量子级联激光器的实时成像研究进展;谭智勇等;《中国光学》;20170215;第68-76页 *

Also Published As

Publication number Publication date
CN111539967A (en) 2020-08-14

Similar Documents

Publication Publication Date Title
CN111539967B (en) Method and system for identifying and processing interference fringe region in terahertz imaging of focal plane
Hou et al. A novel dark channel prior guided variational framework for underwater image restoration
US20210049744A1 (en) Method for image dehazing based on adaptively improved linear global atmospheric light of dark channel
CN112837233B (en) Polarization image defogging method for acquiring transmissivity based on differential polarization
CN115082361B (en) Turbid water body image enhancement method based on image processing
CN105046653B (en) A kind of video raindrop minimizing technology and system
CN101079149A (en) Noise-possessing movement fuzzy image restoration method based on radial basis nerve network
CN116823686A (en) Night infrared and visible light image fusion method based on image enhancement
CN113379640B (en) Multi-stage filtering image denoising method integrating edge information
CN111611907A (en) An infrared target detection method with image enhancement
CN110443166A (en) A kind of licence plate recognition method of haze weather
CN110223344A (en) A kind of infrared small target detection method based on morphology and vision noticing mechanism
CN114612384B (en) Method and system for detecting defects of appearance material of sport protector
CN105913391B (en) A kind of defogging method can be changed Morphological Reconstruction based on shape
CN114359414A (en) Lens smudge identification method and device, computer equipment and storage medium
CN106778631B (en) A design method of heterogeneous iris classifier for fast filtering of fake irises during iris recognition preprocessing
CN111260588A (en) Image enhancement method for high-definition digital CMOS imaging assembly
CN116797484A (en) Image defogging method and system based on improved dark channel prior
CN112270690B (en) Self-adaptive night lane line detection method based on improved CLAHE and sliding window search
CN116485706A (en) Wafer surface defect identification method based on defocused image, storage medium and electronic equipment
CN115035311A (en) Carrier roller detection method based on fusion of visible light and thermal infrared
Yang et al. Infrared image enhancement algorithm based on improved wavelet threshold function and weighted guided filtering
TW202217741A (en) Image processing method
CN113052833A (en) Non-vision field imaging method based on infrared thermal radiation
Ben Gharsallah et al. Computer aided manufacturing method for surface silicon steel inspection based on an efficient anisotropic diffusion algorithm

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant