[go: up one dir, main page]

CN119445050A - A method and device for enhancing image of power drone inspection with adaptive light intensity adjustment - Google Patents

A method and device for enhancing image of power drone inspection with adaptive light intensity adjustment Download PDF

Info

Publication number
CN119445050A
CN119445050A CN202510040432.3A CN202510040432A CN119445050A CN 119445050 A CN119445050 A CN 119445050A CN 202510040432 A CN202510040432 A CN 202510040432A CN 119445050 A CN119445050 A CN 119445050A
Authority
CN
China
Prior art keywords
image
inspection image
pixel
inspection
super
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.)
Granted
Application number
CN202510040432.3A
Other languages
Chinese (zh)
Other versions
CN119445050B (en
Inventor
徐先勇
黄志鸿
肖剑
张文静
陈卓
刘帅
张可人
张辉
钟杭
毛建旭
孔庆宇
张国梁
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.)
Electric Power Research Institute of State Grid Hunan Electric Power Co Ltd
State Grid Hunan Electric Power Co Ltd
Hunan Xiangdian Test Research Institute Co Ltd
State Grid Corp of China SGCC
Original Assignee
Electric Power Research Institute of State Grid Hunan Electric Power Co Ltd
State Grid Hunan Electric Power Co Ltd
Hunan Xiangdian Test Research Institute Co Ltd
State Grid Corp of China SGCC
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 Electric Power Research Institute of State Grid Hunan Electric Power Co Ltd, State Grid Hunan Electric Power Co Ltd, Hunan Xiangdian Test Research Institute Co Ltd, State Grid Corp of China SGCC filed Critical Electric Power Research Institute of State Grid Hunan Electric Power Co Ltd
Priority to CN202510040432.3A priority Critical patent/CN119445050B/en
Publication of CN119445050A publication Critical patent/CN119445050A/en
Application granted granted Critical
Publication of CN119445050B publication Critical patent/CN119445050B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/20Image preprocessing
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06Energy or water supply
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/20Image preprocessing
    • G06V10/26Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G06V10/267Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/40Extraction of image or video features
    • G06V10/42Global feature extraction by analysis of the whole pattern, e.g. using frequency domain transformations or autocorrelation
    • G06V10/431Frequency domain transformation; Autocorrelation
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/70Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/77Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
    • G06V10/80Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level
    • G06V10/809Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level of classification results, e.g. where the classifiers operate on the same input data
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00Scenes; Scene-specific elements
    • G06V20/10Terrestrial scenes
    • G06V20/17Terrestrial scenes taken from planes or by drones

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Business, Economics & Management (AREA)
  • Health & Medical Sciences (AREA)
  • Economics (AREA)
  • General Health & Medical Sciences (AREA)
  • Remote Sensing (AREA)
  • Marketing (AREA)
  • Water Supply & Treatment (AREA)
  • General Business, Economics & Management (AREA)
  • Public Health (AREA)
  • Strategic Management (AREA)
  • Primary Health Care (AREA)
  • Tourism & Hospitality (AREA)
  • Human Resources & Organizations (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)

Abstract

The invention provides a method and a device for enhancing an inspection image of an electric unmanned aerial vehicle with illumination brightness self-adaptive adjustment, wherein the method comprises the steps of carrying out super-pixel segmentation on an acquired inspection image to obtain N super-pixel areas; the method comprises the steps of carrying out intrinsic image decomposition on a super-pixel region to obtain spectral reflection components of the super-pixel region, combining the spectral reflection components of each super-pixel region to obtain an image primary enhancement result, adopting a clear channel feature extraction model to process a patrol image to obtain a bright significance map, and carrying out decision fusion on the image primary enhancement result and the bright significance map to obtain a final illumination brightness enhanced recovery image. The invention can ensure that the electric unmanned aerial vehicle can acquire high-quality inspection images even under extreme illumination conditions (such as backlight and weak light), improves the intelligent operation and inspection safety and reliability of electric power facilities, can be widely applied to the fields of inspection of the electric unmanned aerial vehicle, and has wide application prospect and market value.

Description

Method and device for enhancing inspection image of electric unmanned aerial vehicle with illumination brightness adaptively adjusted
Technical Field
The invention relates to the technical field of image processing, in particular to an electric unmanned aerial vehicle inspection image enhancement method and device capable of adaptively adjusting illumination brightness.
Background
The electric power unmanned aerial vehicle plays an important role in the electric power overhead line inspection task by virtue of the operation convenience and the high-efficiency operation capability. Especially in places (such as mountain areas) which are difficult to reach such as complicated topography, unmanned aerial vehicle can effectively cross natural obstacle, realizes the comprehensive control to transmission line. However, in severe illumination conditions, such as backlight or insufficient light, the quality of the inspection image photographed by the unmanned aerial vehicle may be reduced, so that key features in the image are blurred and difficult to identify.
At this time, the application of the image enhancement technique is particularly important. The technology can obviously improve the image quality and make details in the image clearer and more discernable. By adopting the image enhancement technology, the electric unmanned aerial vehicle can provide accurate and reliable inspection information even under unfavorable illumination conditions, and ensures safe and stable operation and effective maintenance of electric facilities.
In view of the foregoing, there is an urgent need for a method and apparatus for enhancing inspection images of an electric unmanned aerial vehicle with adaptive adjustment of illumination brightness, so as to solve the problems in the prior art.
Disclosure of Invention
The invention aims to provide an electric unmanned aerial vehicle inspection image enhancement method with self-adaptive adjustment of illumination brightness, which aims to inhibit interference of backlight and weak light phenomena on high-quality inspection image imaging and meets the high-quality imaging requirement on the electric inspection image, and the specific technical scheme is as follows:
An electric unmanned aerial vehicle inspection image enhancement method with illumination brightness self-adaptive adjustment comprises the following steps:
performing super-pixel segmentation on the acquired inspection image Y to obtain N super-pixel areas;
The spectral reflection components of each super pixel region are combined to obtain an image primary enhancement result ;
Processing the inspection image Y by adopting a clear channel feature extraction model to obtain a bright significance map;
Preliminary enhancement of image resultsAnd a bright saliency mapDecision fusion is carried out to obtain a restored image with enhanced final illumination brightness
In the above technical solution, preferably, the inspection image Y is represented by a super-pixel area as follows:
(1),
Wherein N is the number of super pixel areas, Represent the firstAnd super pixel areas.
In the above technical solution, it is preferable that the super pixel region is mapped according to formula (3)Performing intrinsic image decomposition:
(3),
Wherein, An nth pixel representing an ith super pixel area in the inspection image Y,Representing the spectral reflectance component of the nth pixel of the ith superpixel region,The luminance component of the nth pixel of the ith super pixel area is represented.
In the above technical solution, it is preferable that the method of formula (3) is performed according to formula (4)AndAnd (3) carrying out solving:
(4),
Wherein, Representation solutionAndIs a target function of (2); Representing the pixel c as the center and the size as A local window of the individual pixels is provided,Parameters for controlling the size of the local window; Representing the similarity of the spectral angles between pixel c and pixel v;
Obtaining a super pixel region according to the formula (3) and the formula (4) Spectral reflection components of each pixel in the array are finally obtained to obtain the whole super-pixel areaIs a spectral reflectance component of (2)
In the above technical solution, preferably, the spectral reflection components of each super pixel region are combined to generate the image preliminary enhancement result:
(6),
Wherein, The spectral reflection component of the ith super pixel area is represented, and N is the number of super pixel areas.
In the above technical scheme, the bright significance map is preferably obtainedThe specific mode of (a) is as follows:
Extracting clear channel characteristics from the inspection image Y through a clear channel characteristic extraction model Channel characteristics are clarifiedAverage fusion is carried out to obtain a mapFor the atlasFourier transform to generate amplitude spectrumFor amplitude spectrumLog calculation to generate log spectrumAnd based on logarithmic spectrumGenerating an average spectrumFinally according to the logarithmic spectrumAnd average spectrumObtaining a bright saliency map of the dark-removed component
In the above technical solution, preferably, the clear channel feature extraction model is:
(7),
Channel characteristics of contrast Average fusion is carried out to obtain a mapThe method comprises the following steps:
(8),
For the atlas Fourier transform to generate amplitude spectrumThe method comprises the following steps:
(9),
Based on log-spectrum Generating an average spectrumThe method comprises the following steps:
(10),
From log spectra And average spectrumObtaining a bright saliency map of the dark-removed componentThe method comprises the following steps:
(12),
Wherein, In order to patrol the red channel of the image,For the green channel of the inspection image,In order to patrol the blue channel of the image,To examine the HUE component of the image in HUE-saturation-intensity variations,To patrol the image for the saturation component in HUE-saturation-intensity variations,For the intensity component of the inspection image in HUE-saturation-intensity variations, p is the number of channels of the bright channel feature M,Representation pairA function of the fourier transform is performed,For a local kxk mean filter feature extractor, k represents the parameters of the mean filter feature extractor,Is an inverse fourier transform operation.
In the above technical solution, preferably, the local kxk mean value filtering feature extractorExpressed as:
(11)。
in the above technical solution, it is preferable that the image is preliminarily enhanced according to formula (13) And a bright saliency mapDecision fusion is carried out to obtain a restored image with enhanced final illumination brightness:
(13)。
The invention also provides an electric unmanned aerial vehicle inspection image enhancement device with the self-adaptive adjustment of the illumination brightness, which comprises a memory and a processor, wherein the processor executes the electric unmanned aerial vehicle inspection image enhancement method with the self-adaptive adjustment of the illumination brightness when running a computer instruction stored in the memory.
The technical scheme of the invention has the following beneficial effects:
The method comprises the steps of firstly, fully mining spatial structure similarity information of an image and intrinsic image spectral reflection information through a super-pixel segmentation and intrinsic image decomposition algorithm to obtain an image primary enhancement result, secondly, mining saliency information of image brightness to generate a bright saliency map, finally, carrying out decision fusion on the image primary enhancement result and the bright saliency map, and jointly mining the spatial structure similarity information of the image, the intrinsic image spectral reflection information and the saliency information of the image brightness to further improve the image enhancement effect.
The invention can ensure that the electric unmanned aerial vehicle can acquire high-quality inspection images even under extreme illumination conditions (such as backlight and weak light), improves the intelligent operation and inspection safety and reliability of electric power facilities, can be widely applied to the fields of inspection of the electric unmanned aerial vehicle, and has wide application prospect and market value.
In addition to the objects, features and advantages described above, the present invention has other objects, features and advantages. The present invention will be described in further detail with reference to the drawings.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:
FIG. 1 is a flow chart of an electric unmanned aerial vehicle inspection image enhancement method with adaptive adjustment of illumination brightness;
FIG. 2a is a graph of imaging effects of a first inspection scenario without the enhancement method of the present invention;
FIG. 2b is a graph of imaging effects of a first inspection scenario employing the enhancement method of the present invention;
FIG. 3a is a graph of imaging effects of a second inspection scenario without the enhancement method of the present invention;
fig. 3b is an image effect diagram of a second inspection scene using the enhancement method of the present invention.
Detailed Description
The present invention will be described more fully hereinafter in order to facilitate an understanding of the present invention, and preferred embodiments of the present invention are set forth. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.
Examples:
Referring to fig. 1, the embodiment provides a method for enhancing an inspection image of an electric unmanned aerial vehicle with adaptive adjustment of illumination brightness, which comprises the following steps:
Step S1, carrying out super-pixel segmentation on an acquired inspection image to obtain N super-pixel areas;
Preferably, in step S1, the electric power unmanned aerial vehicle is used to photograph the electric power equipment, so as to obtain an original visible light image Y (i.e. a patrol image Y) in jpg format. Specifically, the super-pixel region is a homogeneous region in the inspection image, and pixels in the homogeneous region have similar spatial information. In the embodiment, an entropy rate super-pixel segmentation technology (ERS) is adopted to obtain a super-pixel region in a patrol image, a principal component analysis algorithm is adopted to process the patrol image Y to reserve a first principal component, then graph optimization problem modeling is carried out on the first principal component, and an ERS algorithm is adopted to generate a series of super-pixel edges so as to obtain a plurality of homogeneous regions (namely super-pixel regions).
The entropy rate super pixel segmentation (ERS) is shown in the prior art "Entropy rate superpixel segmentation", Liu M, Tuzel O, Ramalingam S, et al, Proc of Processing IEEE Confense Computer Vision Pattern Recognition, 2011, 2097–2104; and the principal component analysis algorithm is shown in the prior art "Two-stage image denoising by principal component analysis with local pixel grouping", Zhang L, Dong W, Zhang D, et al, Pattern Recognition, 2010, 43(4):1531–1549;, and these two techniques are not further described in this embodiment.
Specifically, the inspection image Y is represented by a super-pixel area as:
(1),
Wherein N is the number of super pixel areas, Represent the firstAnd super pixel areas.
S2, carrying out intrinsic image decomposition on the super-pixel areas to obtain spectral reflection components of the super-pixel areas, combining the spectral reflection components of the super-pixel areas to obtain an image primary enhancement result;
Specifically, the inspection image Y can be decomposed into a spectral reflectance component R and a luminance component S by the intrinsic image, and this process is expressed as:
(2),
Wherein: First representing inspection image A number of pixels of the pixel array are arranged,Representing the first of the inspection imagesThe spectral reflection component of each pixel,Representing the first of the inspection imagesA luminance component of each pixel;
Further, wherein ,,;Respectively represent the first inspection image YThe elements of the individual pixels on the red, green and blue channels,Respectively represent the first of the inspection imagesThe spectral reflection component of each pixel is an element on the red, green and blue channels,Respectively represent the first of the inspection imagesThe luminance component of each pixel is an element on three channels of red, green and blue, r represents the red channel, g represents the green channel, and b represents the blue channel.
Furthermore, because the pixel value of the brightness component of the image is directly related to factors such as scene illumination, shadow and the like and the spectral reflection component is directly related to the material of the object, the brightness component in the image is removed, only the spectral reflection component in the image is reserved, and the interference of extreme illumination in the image can be well restrained. Thus, for super pixel regionThe intrinsic image decomposition is performed as:
(3),
Wherein, An nth pixel representing an ith super pixel area in the inspection image Y,Representing the spectral reflectance component of the nth pixel of the ith superpixel region,The luminance component of the nth pixel of the ith super pixel area is represented.
Further, in formula (3)AndThe method can be obtained by solving the following optimization equation:
(4),
Wherein, Representation solutionAndIs a target function of (2); Representing the pixel c as the center and the size as A local window of the individual pixels is provided,Parameters for controlling the size of the local window; The similarity of the spectral angles between pixel c and pixel v is indicated.
Obtaining a super pixel region according to the formula (3) and the formula (4)Spectral reflection component of each pixel in (B)Finally, the whole super pixel area is obtainedIs a spectral reflectance component of (2):
(5),
Wherein sup represents the super pixel regionThe number of inner pixels.
Combining spectral reflectance components of each super-pixel region to generate an image preliminary enhancement result:
(6),
Wherein, The spectral reflection component of the ith super pixel area is represented, and N is the number of super pixel areas.
S3, processing the inspection image by adopting a clear channel feature extraction model to obtain a bright significance map;
Specifically, the clear channel feature extraction model constructed in this embodiment is:
(7),
Wherein, In order to be able to characterize the channel,In order to patrol the red channel of the image,For the green channel of the inspection image,In order to patrol the blue channel of the image,To examine the HUE component of the image in HUE-saturation-intensity variations,To patrol the image for the saturation component in HUE-saturation-intensity variations,To examine the intensity component of the image in HUE-saturation-intensity variations.
The characteristic wave band of the maximum brightness of the inspection image is extracted to obtain the characteristic of the bright channelChannel characteristics are clarifiedAverage fusion is carried out to realize dimension reduction treatment on the characteristics, thereby obtaining a map:
(8),
Wherein p is the number of channels of the clear channel feature M.
To further mine the mapA significant map with bright image features, dark image features in the image suppressed, and a patternFourier transform to generate amplitude spectrum:
(9),
Wherein, Representation pairA function of fourier transform is performed.
Further, for amplitude spectrumLog calculation to generate log spectrumAnd generates an average spectrum according to formula (10):
(10),
Wherein, Is a local k x k means filter feature extractor.
Preferably, the local kxk means filter feature extractorThe method comprises the following steps:
(11),
Where k is a parameter of the mean filter feature extractor, in this embodiment k=3 is taken.
Further, the average spectrumCan reflect redundant dark area characteristics in an image, and then recover by a formula (12) to obtain a bright significance map of dark-removed components:
(12),
Wherein, Is an inverse fourier transform operation.
Step S4, the image is initially enhancedAnd a bright saliency mapDecision fusion is carried out to obtain a restored image with enhanced final illumination brightness
Specifically, the image is initially enhanced according to equation (13)And a bright saliency mapDecision fusion is carried out, and interference of backlight and weak light phenomena in the image on high-quality inspection image imaging is inhibited:
(13)。
Referring to fig. 2a, fig. 2b, fig. 3a and fig. 3b, wherein fig. 2a and fig. 2b are a set of identical inspection scenes, it can be seen from fig. 2a and fig. 2b that the image is overall dark when the enhancing method of the present embodiment is not adopted, the features in the image are blurred and difficult to identify, the image is overall bright when the enhancing method of the present embodiment is adopted, the features in the image are clearer and easier to identify in comparison, and fig. 3a and fig. 3b are a set of identical inspection scenes, it can be seen from fig. 3a and fig. 3b that the image is overall dark, the features on the lower right side are blurred, the smoke in the image is difficult to identify when the enhancing method of the present embodiment is not adopted, the image is brighter and the features on the lower right side are clearer and easier to identify, and the smoke on the lower right side of the image can be clearly seen. Therefore, the enhancement method of the embodiment can effectively enhance the illumination brightness in the image, even under extreme illumination conditions (such as backlight and weak light), the electric unmanned aerial vehicle can acquire high-quality inspection images, the intelligent operation and inspection safety and reliability of the electric power facility are improved, the enhancement method can be widely applied to the fields of inspection of the electric unmanned aerial vehicle and the like, and the enhancement method has wide application prospects and market values.
The embodiment also provides an electric unmanned aerial vehicle inspection image enhancement device with self-adaptive adjustment of illumination brightness, which comprises a memory and a processor, wherein the processor executes the electric unmanned aerial vehicle inspection image enhancement method with self-adaptive adjustment of illumination brightness when running a computer instruction stored in the memory so as to process an inspection image.
The embodiment also provides another power unmanned aerial vehicle inspection image enhancement device with self-adaptive adjustment of illumination brightness, which comprises an image acquisition unit and a data calculation processing unit, wherein the image acquisition unit is used for shooting power equipment to acquire an inspection image, and the data calculation processing unit processes the acquired inspection image according to the power unmanned aerial vehicle inspection image enhancement method with self-adaptive adjustment of illumination brightness.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (9)

1.一种光照亮度自适应调整的电力无人机巡检图像增强方法,其特征在于,包括:1. A method for enhancing an image of an electric power drone inspection with adaptive light intensity adjustment, comprising: 对采集的巡检图像Y进行超像素分割,得到N个超像素区域;Perform superpixel segmentation on the collected inspection image Y to obtain N superpixel regions; 对超像素区域进行本征图像分解,得到超像素区域的光谱反射成分;将各超像素区域的光谱反射成分组合得到图像初步增强结果Perform intrinsic image decomposition on the super-pixel region to obtain the spectral reflectance component of the super-pixel region; combine the spectral reflectance components of each super-pixel region to obtain the preliminary image enhancement result ; 采用明通道特征提取模型对巡检图像Y进行处理,得到明亮显著性图谱The bright channel feature extraction model is used to process the inspection image Y to obtain the bright saliency map ; 将图像初步增强结果和明亮显著性图谱进行决策融合,得到最终光照亮度增强的恢复图像The image is initially enhanced. and Brightness Saliency Map Perform decision fusion to obtain the final restored image with enhanced light intensity ; 其中,得到明亮显著性图谱的具体方式为:通过明通道特征提取模型对巡检图像Y提取明通道特征,对明通道特征进行平均融合得到图谱,对图谱进行傅里叶变换生成振幅谱,对振幅谱进行对数计算生成对数谱,并基于对数谱生成平均谱,最后根据对数谱和平均谱获得去黑暗成分的明亮显著性图谱Among them, the bright saliency map is obtained The specific method is: extract the bright channel features of the inspection image Y through the bright channel feature extraction model , for bright channel characteristics Perform average fusion to obtain the atlas , for the atlas Perform Fourier transform to generate amplitude spectrum , for the amplitude spectrum Perform logarithmic calculations to generate logarithmic spectra , and based on the logarithmic spectrum Generate average spectrum , and finally according to the logarithmic spectrum and average spectrum Obtain a bright saliency map without dark components . 2.根据权利要求1所述的光照亮度自适应调整的电力无人机巡检图像增强方法,其特征在于,所述巡检图像Y用超像素区域表示为:2. The method for enhancing the inspection image of an electric power drone with adaptive light intensity adjustment according to claim 1, wherein the inspection image Y is represented by a superpixel region as follows: (1), (1), 其中,N为超像素区域的个数,表示第个超像素区域。Where N is the number of superpixel regions, Indicates Superpixel area. 3.根据权利要求1所述的光照亮度自适应调整的电力无人机巡检图像增强方法,其特征在于,根据公式(3)对超像素区域进行本征图像分解:3. The method for enhancing the inspection image of an electric drone with adaptive light intensity adjustment according to claim 1 is characterized in that the superpixel area is Perform intrinsic image decomposition: (3), (3), 其中,表示巡检图像Y中第i个超像素区域的第n个像素,表示第i个超像素区域的第n个像素的光谱反射成分,表示第i个超像素区域的第n个像素的亮度成分。in, represents the nth pixel in the i- th superpixel region in the inspection image Y , represents the spectral reflectance component of the nth pixel in the i - th superpixel area, Represents the brightness component of the nth pixel in the i- th superpixel area. 4.根据权利要求3所述的光照亮度自适应调整的电力无人机巡检图像增强方法,其特征在于,根据公式(4)对公式(3)中的进行求解:4. The method for enhancing the inspection image of an electric drone with adaptive light intensity adjustment according to claim 3 is characterized in that: according to formula (4), and To solve: (4), (4), 其中,表示求解的目标函数;表示以像素c为中心、大小为个像素的局部窗口,为控制局部窗口大小的参数;表示像素c和像素v间光谱角的相似度;in, Indicates solution and The objective function of It means that the pixel c is the center and the size is A local window of pixels, Parameters to control the size of the local window; Represents the similarity of the spectral angles between pixel c and pixel v ; 根据公式(3)和公式(4)获取超像素区域中各像素的光谱反射成分,最终得到整个超像素区域的光谱反射成分According to formula (3) and formula (4), the super pixel area is obtained The spectral reflectance component of each pixel in the final super-pixel area is obtained The spectral reflectance component . 5.根据权利要求4所述的光照亮度自适应调整的电力无人机巡检图像增强方法,其特征在于,将每个超像素区域的光谱反射成分进行组合生成图像初步增强结果5. The method for enhancing the inspection image of an electric power drone with adaptive light intensity adjustment according to claim 4 is characterized in that the spectral reflectance components of each superpixel area are combined to generate a preliminary image enhancement result. : (6), (6), 其中,表示第i个超像素区域的光谱反射成分,N为超像素区域的个数。in, represents the spectral reflectance component of the i -th superpixel area, and N is the number of superpixel areas. 6.根据权利要求1所述的光照亮度自适应调整的电力无人机巡检图像增强方法,其特征在于:6. The method for enhancing the inspection image of an electric power drone with adaptive light intensity adjustment according to claim 1 is characterized in that: 所述明通道特征提取模型为:The bright channel feature extraction model is: (7), (7), 对明通道特征进行平均融合得到图谱为:Bright channel characteristics Perform average fusion to obtain the atlas for: (8), (8), 对图谱进行傅里叶变换生成振幅谱为:Pair Spectrum Perform Fourier transform to generate amplitude spectrum for: (9), (9), 基于对数谱生成平均谱为:Based on the logarithmic spectrum Generate average spectrum for: (10), (10), 根据对数谱和平均谱获得去黑暗成分的明亮显著性图谱为:According to the logarithmic spectrum and average spectrum Obtain a bright saliency map without dark components for: (12), (12), 其中,为巡检图像的红色通道,为巡检图像的绿色通道,为巡检图像的蓝色通道,为巡检图像在HUE色调-饱和度-强度变化中的色调分量,为巡检图像在HUE色调-饱和度-强度变化中的饱和度分量,为巡检图像在HUE色调-饱和度-强度变化中的强度分量,p为明通道特征M的通道个数,表示对进行傅里叶变换的函数,为局部k×k均值滤波特征提取器,k表示均值滤波特征提取器的参数,为傅里叶逆变换操作。in, is the red channel of the inspection image, is the green channel of the inspection image, is the blue channel of the inspection image, is the hue component of the inspection image in the HUE hue-saturation-intensity change, is the saturation component of the inspection image in the HUE hue-saturation-intensity change, is the intensity component of the inspection image in the HUE hue-saturation-intensity change, p is the number of channels of the bright channel feature M , Express The function to perform Fourier transform, is a local k×k mean filter feature extractor, k represents the parameter of the mean filter feature extractor, is the inverse Fourier transform operation. 7.根据权利要求6所述的光照亮度自适应调整的电力无人机巡检图像增强方法,其特征在于,所述局部k×k均值滤波特征提取器表示为:7. The method for enhancing the inspection image of an electric power drone with adaptive light intensity adjustment according to claim 6 is characterized in that the local k×k mean filter feature extractor It is expressed as: (11)。 (11). 8.根据权利要求1所述的光照亮度自适应调整的电力无人机巡检图像增强方法,其特征在于,根据公式(13)将图像初步增强结果和明亮显著性图谱进行决策融合,得到最终光照亮度增强的恢复图像8. The method for enhancing the inspection image of an electric power drone with adaptive light intensity adjustment according to claim 1 is characterized in that the image preliminary enhancement result is converted into and Brightness Saliency Map Perform decision fusion to obtain the final restored image with enhanced light intensity : (13)。 (13). 9.一种光照亮度自适应调整的电力无人机巡检图像增强装置,其特征在于,包括存储器和处理器,所述处理器运行存储器中存储的计算机指令时,执行如权利要求1-8任意一项所述的光照亮度自适应调整的电力无人机巡检图像增强方法。9. An electric power drone inspection image enhancement device with adaptive light intensity adjustment, characterized in that it comprises a memory and a processor, and when the processor runs the computer instructions stored in the memory, it executes the electric power drone inspection image enhancement method with adaptive light intensity adjustment as described in any one of claims 1-8.
CN202510040432.3A 2025-01-10 2025-01-10 Method and device for enhancing inspection image of electric unmanned aerial vehicle with illumination brightness adaptively adjusted Active CN119445050B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202510040432.3A CN119445050B (en) 2025-01-10 2025-01-10 Method and device for enhancing inspection image of electric unmanned aerial vehicle with illumination brightness adaptively adjusted

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202510040432.3A CN119445050B (en) 2025-01-10 2025-01-10 Method and device for enhancing inspection image of electric unmanned aerial vehicle with illumination brightness adaptively adjusted

Publications (2)

Publication Number Publication Date
CN119445050A true CN119445050A (en) 2025-02-14
CN119445050B CN119445050B (en) 2025-04-15

Family

ID=94529236

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202510040432.3A Active CN119445050B (en) 2025-01-10 2025-01-10 Method and device for enhancing inspection image of electric unmanned aerial vehicle with illumination brightness adaptively adjusted

Country Status (1)

Country Link
CN (1) CN119445050B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7580586B1 (en) * 2005-04-19 2009-08-25 Lockheed Martin Corporation Enhanced recovery of low spatial frequency spectral information in a Fizeau Fourier transform spectrometer
CN103903229A (en) * 2014-03-13 2014-07-02 中安消技术有限公司 Night image enhancement method and device
US20220044375A1 (en) * 2019-12-17 2022-02-10 Dalian University Of Technology Saliency Map Enhancement-Based Infrared and Visible Light Fusion Method
CN117274079A (en) * 2023-09-05 2023-12-22 西南交通大学 Low-brightness image enhancement system and method thereof
CN118212150A (en) * 2024-02-27 2024-06-18 智洋创新科技股份有限公司 Image enhancement method, system and storage medium for power transmission channel
CN118864266A (en) * 2024-06-28 2024-10-29 北京航空航天大学 Lightweight fusion method of infrared and visible light images based on spatial and frequency domain information

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7580586B1 (en) * 2005-04-19 2009-08-25 Lockheed Martin Corporation Enhanced recovery of low spatial frequency spectral information in a Fizeau Fourier transform spectrometer
CN103903229A (en) * 2014-03-13 2014-07-02 中安消技术有限公司 Night image enhancement method and device
US20220044375A1 (en) * 2019-12-17 2022-02-10 Dalian University Of Technology Saliency Map Enhancement-Based Infrared and Visible Light Fusion Method
CN117274079A (en) * 2023-09-05 2023-12-22 西南交通大学 Low-brightness image enhancement system and method thereof
CN118212150A (en) * 2024-02-27 2024-06-18 智洋创新科技股份有限公司 Image enhancement method, system and storage medium for power transmission channel
CN118864266A (en) * 2024-06-28 2024-10-29 北京航空航天大学 Lightweight fusion method of infrared and visible light images based on spatial and frequency domain information

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
JUNHAO TAN等: "Wavelet-Based Mamba with Fourier Adjustment for Low-Light Image Enhancement", 《ARXIV.ORG》, 31 October 2024 (2024-10-31), pages 1 - 16 *
廖常青: ""结构保持的图像增强研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》, no. 05, 15 May 2024 (2024-05-15), pages 138 - 896 *
张文杰: "雾霾图像去雾技术与应用研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》, no. 05, 31 May 2024 (2024-05-31), pages 138 - 1440 *
谭海曙;周富强;熊瑛;李学夔;: "基于神经网络的图像亮度和对比度自适应增强", 光电子.激光, no. 12, 15 December 2010 (2010-12-15), pages 141 - 144 *

Also Published As

Publication number Publication date
CN119445050B (en) 2025-04-15

Similar Documents

Publication Publication Date Title
CN113658057B (en) Swin converter low-light-level image enhancement method
EP4050558B1 (en) Image fusion method and apparatus, storage medium, and electronic device
EP4109392A1 (en) Image processing method and image processing device
CN111784605B (en) Image noise reduction method based on region guidance, computer device and computer readable storage medium
WO2022042049A1 (en) Image fusion method, and training method and apparatus for image fusion model
CN106952246A (en) Visible-infrared image enhancement color fusion method based on visual attention characteristics
CN111429370A (en) Method and system for enhancing images in coal mine and computer storage medium
CN112712481A (en) Structure-texture sensing method aiming at low-light image enhancement
Kaur et al. Single image dehazing with dark channel prior
CN110349114A (en) Applied to the image enchancing method of AOI equipment, device and road video monitoring equipment
CN114331937B (en) Multi-source image fusion method based on feedback iterative adjustment under low illumination condition
CN112822343A (en) Night video oriented sharpening method and storage medium
CN104933728A (en) Mixed motion target detection method
CN111476732A (en) A method and system for image fusion and denoising
CN117994146A (en) A method for infrared and low-light image fusion based on non-subsampled shearing transform
CN105405110A (en) Uneven light compensation method
CN118334481A (en) Infrared and low-light visible light image fusion method, storage medium, and electronic device
CN115035393B (en) Stroboscopic scene classification method, model training method, related device and electronic equipment
CN119445050B (en) Method and device for enhancing inspection image of electric unmanned aerial vehicle with illumination brightness adaptively adjusted
CN114782404A (en) A computer image processing device
CN113781329B (en) Fog removing method for remote sensing image
CN117152404B (en) Low-light image enhancement method for space self-adaptive supervised learning
CN116934642A (en) Image processing method, computer device, readable medium and video monitoring system
CN111383242B (en) Image fog penetration processing method and device
Wang et al. LLD-GAN: An end-to-end network for low-light image demosaicking

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