WO2021217445A1 - 图像处理方法、装置、系统和存储介质 - Google Patents

图像处理方法、装置、系统和存储介质 Download PDF

Info

Publication number
WO2021217445A1
WO2021217445A1 PCT/CN2020/087574 CN2020087574W WO2021217445A1 WO 2021217445 A1 WO2021217445 A1 WO 2021217445A1 CN 2020087574 W CN2020087574 W CN 2020087574W WO 2021217445 A1 WO2021217445 A1 WO 2021217445A1
Authority
WO
WIPO (PCT)
Prior art keywords
stretched
stretching
area
region
image
Prior art date
Application number
PCT/CN2020/087574
Other languages
English (en)
French (fr)
Inventor
张青涛
庹伟
赵新涛
Original Assignee
深圳市大疆创新科技有限公司
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 深圳市大疆创新科技有限公司 filed Critical 深圳市大疆创新科技有限公司
Priority to PCT/CN2020/087574 priority Critical patent/WO2021217445A1/zh
Publication of WO2021217445A1 publication Critical patent/WO2021217445A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
    • H04N21/44—Processing of video elementary streams, e.g. splicing a video clip retrieved from local storage with an incoming video stream or rendering scenes according to encoded video stream scene graphs

Definitions

  • This application relates to the field of image processing technology, and in particular to an image processing method, device, system and storage medium.
  • Infrared image is an image formed by obtaining the infrared light intensity of an object. It has some problems that affect image quality, such as low contrast, concentrated gray histogram, and obvious noise jump. Therefore, it is often necessary to process infrared images.
  • the infrared image is subjected to contrast stretching processing according to predetermined gray-scale stretching parameters.
  • the contrast stretching processing performed on the infrared image has the problem that the gray-scale stretching parameters may not be applicable to some local areas of the infrared image, which leads to the ineffective processing of the infrared image.
  • the problem is that the image quality is damaged.
  • the present application provides an image processing method, device, system, and storage medium, which are used to solve the problems of poor infrared image processing effect and impaired image quality existing in the existing image processing technology.
  • the present application provides an image processing method for infrared image processing, and the method includes:
  • the initial stretching parameter is adjusted to obtain the controlled stretching parameter
  • the stretch processing is performed on the region to be stretched.
  • the present application provides an image processing device, the device is used for infrared image processing, the device includes a processor and a memory, the processor is in communication connection with the memory, and the memory stores the Instructions executed by the processor, the processor executes the instructions to perform the following steps:
  • the initial stretching parameter is adjusted to obtain the controlled stretching parameter
  • the stretch processing is performed on the region to be stretched.
  • the present application provides an image processing system, including an infrared camera and the image processing device according to the second aspect, wherein the infrared camera is correspondingly connected to the image processing device, and the image processing device is used for The infrared image collected by the infrared camera is processed.
  • the present application provides a computer-readable storage medium on which a computer program is stored, and the computer program is executed by a processor to implement any of the above methods.
  • the present application provides an image processing method, device, equipment, and storage medium.
  • the region to be stretched corresponds to at least a part of the input image;
  • the gray-scale feature of the stretched area determines the initial stretch parameter of the area to be stretched; according to the stretch control feature, the initial stretch parameter is regulated to obtain the regulated stretch parameter; Stretching parameters, stretching the region to be stretched.
  • the initial stretching parameters are determined according to the gray-scale characteristics of the area to be stretched
  • the initial stretching parameters are adjusted according to the stretching control characteristics to obtain the adjusted stretching parameters
  • the gray-scale of the area to be stretched is performed according to the adjusted stretching parameters Stretching processing, the adjustment of the stretch parameters and the area to be stretched has a high degree of matching, strong applicability, high image quality of the processed infrared image, and significant image enhancement effect.
  • FIG. 1 is a schematic flowchart of an image processing method provided by an embodiment of this application.
  • Fig. 1a is a schematic diagram of determining initial stretching parameters according to the gray value range of the area to be stretched according to an embodiment of the application;
  • Figure 1b is a schematic diagram of determining the gray-scale probability density function according to the gray-scale histogram of the area to be stretched according to an embodiment of the application;
  • FIG. 2 is a schematic flowchart of another image processing method provided by an embodiment of this application.
  • FIG. 2a is a schematic flowchart of still another image processing method provided by an embodiment of this application.
  • FIG. 3 is a schematic structural diagram of an image processing device provided by an embodiment of the application.
  • FIG. 4 is a schematic structural diagram of an image processing system provided by an embodiment of the application.
  • Signal-Noise Ratio is the ratio of the signal mean value to the background signal value variance.
  • the signal mean value often refers to the gray value, and the physical meaning of the background signal value variance is noise power.
  • an infrared image is an image formed by acquiring the infrared light intensity of an object.
  • the straight-out image of the infrared sensor has some effects such as low contrast, concentrated grayscale histogram, and obvious noise jump. Image quality issues.
  • the infrared images collected by the infrared camera set on the drone have the problems of low signal-to-noise ratio, insufficient temperature display, excessive dark areas, and overexposed bright areas, which cannot display objects in the scene.
  • the infrared image enhancement processing method commonly used in the prior art includes: performing a global area contrast stretching process on the infrared image according to predetermined gray-scale stretching parameters.
  • the predetermined gray scale stretching parameters may not be applicable to some local areas of the infrared image, for example, the predetermined gray scale
  • the stretching parameters are not suitable for the corner areas of the infrared image, resulting in brighter corners of the infrared image after the gray-scale stretching process, or the predetermined gray-scale stretching parameters are not suitable for the noise area of the infrared image.
  • the noise area and the normal area use the same gray-scale stretching parameter for gray-scale stretching, which leads to higher gray values in the noise area and more obvious image flicker. These all lead to poor enhancement of the infrared image and impaired image quality. problem.
  • This application proposes an image processing method, device, system and storage medium, aiming to solve the above technical problems.
  • Fig. 1 is a schematic flowchart of an image processing method provided by an embodiment of the application. This method is used for infrared image processing. As shown in Fig. 1, the method includes:
  • Step 101 Determine the grayscale feature and the stretching control feature of the area to be stretched, and the area to be stretched corresponds to at least a part of the input image.
  • the execution subject of this embodiment is an image processing module, and the image processing module is set in a control terminal or a movable platform.
  • the movable platform includes unmanned aerial vehicles, unmanned vehicles, and movable robots.
  • the execution subject is the image processing module provided in the drone as an example for description.
  • the area to be stretched is an area in the input image that needs to be stretched by gray scale, which corresponds to at least a part of the input image, and specifically may be the entire area of the input image, or may be a part of the input image.
  • the area to be stretched can be an entire input image, or a partial area of the input image, or an area corresponding to an object in the input image, or an area corresponding to a scene, or a certain temperature range
  • the corresponding area can also be the area of interest specified by the user in the input image, or it can be the boundary area of the two scenes in the input image.
  • the gray-scale feature can reflect the gray-scale distribution of the pixels of the image, and can specifically include features such as gray-scale value, gray-scale value range, gray-scale histogram, and gray-scale probability density function.
  • the maximum gray level range of the pixels of the image is [0,255]. When the gray value is 0, the corresponding color is white; when the gray value is 255, the corresponding color is black.
  • the stretch control feature characterizes the feature used to control the stretch parameter, including but not limited to the noise of the input image, the signal-to-noise ratio, or at least part of the edge of the region to be stretched.
  • Infrared images collected by infrared cameras or infrared collectors usually include noise.
  • Noise can include any one or more of single-point random noise, horizontal stripe noise, vertical stripe noise, and thermal background noise. Noise affects the image quality of infrared images. , It is easy to cause image artifacts.
  • the signal-to-noise ratio of the input image is the ratio of the pixel gray value to the noise power.
  • At least part of the edges of the region to be stretched can be detected by edge detection operators.
  • the edge detection operators include Roberts Cross operator, Prewitt operator, Sobel operator, Canny operator, Marr-Hildreth operator, and so on.
  • Step 102 Determine the initial stretching parameters of the area to be stretched according to the gray-scale characteristics of the area to be stretched.
  • the stretching parameters may include stretching strength, grayscale offset amplitude, etc., according to the grayscale characteristics of the region to be stretched, the initial stretching parameters of the region to be stretched are determined, including at least one of the following Species: According to the gray value range of the area to be stretched, determine the initial stretching parameters of the area to be stretched; according to the mapping relationship between the gray value and the pixel probability density in the global area of the area to be stretched, determine the to be stretched The initial stretching parameters of the region; according to the mapping relationship between the gray value and the pixel probability density in the part of the region to be stretched, the local stretching parameters of the region to be stretched are determined, and the local stretching parameters constitute the initial stretching parameters. Among them, the local stretch parameters correspond to some areas.
  • the initial stretching parameters of the area to be stretched are determined.
  • FIG. The schematic diagram of the parameters, as shown in Figure 1a, f (x, y) represents the gray value of a certain pixel (x, y) in the area to be stretched, and g (x, y) represents the same pixel in the area to be stretched.
  • the gray value of the point (x, y) after being stretched, the gray value range of the area to be stretched is [a, b], and the gray value range of the expected stretch is [m, n], where (nm )>(ba), the stretching process can be expressed as follows:
  • (n-m)/(b-a) is the initial stretching parameter determined according to the gray value range of the area to be stretched.
  • the initial stretching parameter of the area to be stretched is determined.
  • the mapping relationship between the gray value and the pixel probability density may be the pixel probability density function of the gray value or the cumulative distribution function of the gray value.
  • the mapping relationship between the gray value and the pixel probability density is a cumulative distribution function of the gray value.
  • the grayscale histogram of the global area of the region to be stretched may be obtained first, and then the probability density function of the grayscale value may be determined according to the grayscale histogram, and then the grayscale value may be determined according to the probability density function of the grayscale value.
  • Fig. 1b is a schematic diagram of determining the gray-scale probability density function according to the gray-scale histogram of the area to be stretched according to the present embodiment.
  • the gray-scale statistical histogram is A discrete function, assuming that the total number of pixels in the area to be stretched is N and the number of gray values it has is L, the histogram can be expressed as:
  • i takes a value within the range of gray value levels [0, L-1]
  • r i represents the gray value corresponding to the i-th gray value level
  • n i represents the gray value r
  • the probability density function of gray value is determined as:
  • p (r i ) r i represents the probability that the gradation value appears. Integrating the probability density function of the gray value shown in formula (3) can obtain the cumulative distribution function of the gray value, and then determine the initial stretching parameter of the region to be stretched according to the cumulative distribution function of the gray value.
  • the local stretch parameter of the area to be stretched is determined.
  • the local stretch parameter constitutes the initial stretch parameter, where the local stretch parameter and Corresponding to some areas.
  • the above partial area is obtained by dividing the global area of the area to be stretched.
  • the global area of the area to be stretched can be divided into 4*4, 8*8, 16*16, 32*32 or 64 *64 and other specifications are divided into blocks to obtain multiple blocks of the above partial areas.
  • each partial area according to the mapping relationship between the gray value and the pixel probability density, multiple local stretching parameters corresponding to each partial area in the area to be stretched are determined, and the multiple local stretching parameters constitute the to-be-stretched area.
  • the initial stretch parameters of the stretched area In each partial area, the method and principle of determining the local stretching parameter corresponding to the partial area according to the mapping relationship between the gray value and the pixel probability density is the same as the aforementioned method and principle according to the gray value in the global area of the area to be stretched.
  • the method and principle of the mapping relationship between the pixel probability density and the pixel probability density to determine the initial stretching parameters of the region to be stretched are similar or the same.
  • the above three methods are only used for the explanation of the present embodiment, and are not used to limit the present application.
  • the present application may also use other methods to determine the initial stretching parameters of the region to be stretched, which will not be repeated in this embodiment.
  • Step 103 According to the characteristics of the stretching control, the initial stretching parameters are adjusted to obtain the adjusted stretching parameters.
  • the stretch control feature includes the noise of the input image, the signal-to-noise ratio, or at least part of the edge of the region to be stretched.
  • the initial stretching parameters obtained in step 102 only consider the influence of the gray-scale characteristics of the area to be stretched on the stretching parameters. However, in the actual stretching process, in order to ensure that better image quality can be obtained after stretching, When determining the stretching parameters, it is also necessary to consider the influence of other factors, such as the noise of the infrared image of the area to be stretched, the signal-to-noise ratio, or at least part of the edge of the area to be stretched.
  • the stretching control feature is used to adjust the initial stretching parameters obtained in step 102 to obtain stretching parameters with better stretching effects.
  • the initial stretching parameters can be adjusted.
  • the initial stretching parameters can be adjusted according to the noise intensity before stretching, so that the noise intensity after stretching is small, or the noise intensity after stretching is low.
  • the signal-to-noise ratio of the infrared image is large, thereby improving the image quality of the stretched infrared image;
  • the initial stretching parameters can be adjusted according to the gray value of at least part of the edge of the area to be stretched to make the stretch
  • the image quality of the center part and the corner part of the stretched infrared image is similar or the same.
  • Step 104 Perform a stretching process on the area to be stretched according to the adjusted stretching parameters.
  • the gray value of the area to be stretched is stretched according to the control and stretch parameters, and the gray value of the stretched area is stretched according to the control and stretch parameters.
  • This method can be used.
  • the conventional gray-scale value stretching method in the field for example, can use the algorithm program suppressed in the field to perform gray-scale value stretching processing on the area to be stretched, which will not be repeated here in this embodiment.
  • the area to be stretched corresponds to at least a part of the input image; according to the gray-scale characteristics of the area to be stretched, the initial of the area to be stretched is determined Stretching parameters: According to the characteristics of the stretching control, the initial stretching parameters are adjusted to obtain the adjusted stretching parameters; according to the adjusted stretching parameters, the stretched area to be stretched is processed.
  • the initial stretching parameters are determined according to the gray-scale characteristics of the area to be stretched
  • the initial stretching parameters are adjusted according to the stretching control characteristics to obtain the adjusted stretching parameters
  • the gray-scale of the area to be stretched is performed according to the adjusted stretching parameters Stretching processing, the adjustment of the stretch parameters and the area to be stretched has a high degree of matching, strong applicability, high image quality of the processed infrared image, and significant image enhancement effect.
  • Fig. 2 is a schematic flowchart of another image processing method provided by an embodiment of the application. Based on Fig. 1, as shown in Fig. 2, the method includes:
  • Step 201 Determine the area to be stretched.
  • the area to be stretched is an area in the input image that needs to be grayscale stretched, which corresponds to at least a part of the input image.
  • the area to be stretched may be a global area of the input image, or Part of the area obtained after the input image is segmented, or the area corresponding to an object in the input image, or the area corresponding to a certain scene, or the area corresponding to a certain temperature interval, or the area of interest specified by the user, etc. .
  • the number of regions to be stretched is at least one.
  • the area to be stretched can be obtained through a machine learning method. Specifically, when the image effect of an object (such as a person or other scene) in the infrared image needs to be enhanced , The input image can be recognized through a machine learning algorithm, and the area where the object to be enhanced is located is identified as the area to be stretched. It is also possible to determine the area to be stretched according to the input control instruction, or to determine the area to be stretched according to a preset program.
  • the control instruction may be input by the user before the area to be stretched is determined, and the preset program may be preset in the memory, or may be preset by the user before the area to be stretched is determined.
  • the input control instruction or the preset program is used to indicate the rules or requirements for determining the area to be stretched.
  • the input control instruction includes at least one of the following: dividing the area to be stretched according to the image temperature; according to the objects in the image Divide the area to be stretched; divide the area to be stretched according to the scene in the image; divide the specified area in the input image into the area to be stretched. This part of the area can be the user's area of interest or the control command input by the user
  • the indicated arbitrary area may be, for example, the upper half area or the lower right quarter area of the image.
  • the image area corresponding to a certain target temperature interval may be used as the area to be stretched, or different image areas corresponding to different temperature intervals may be used as multiple areas to be stretched.
  • the image area corresponding to a certain target temperature range is used as the area to be stretched, the area to be stretched can be stretched with a higher tensile strength according to user needs, and the image area corresponding to other temperature ranges can be stretched with a lower stretch.
  • the intensity is stretched, or not stretched, so as to improve the contrast of the stretched image corresponding to the target temperature range.
  • different stretch parameters can be used to stretch the regions to be stretched corresponding to different temperature intervals to improve the stretched image Contrast, for example, for the area to be stretched corresponding to the ultra-low temperature interval that the user is not interested in, a lower tensile strength can be used for stretching; for the area to be stretched corresponding to the low temperature interval or the medium temperature interval, it can be appropriately increased compared to the ultra-low temperature interval Stretching is performed after the tensile strength; for the area to be stretched corresponding to the high temperature range, a higher tensile strength can be used for stretching according to different scenarios and/or user needs.
  • the method and principle of gray-scale stretching of the regions to be stretched corresponding to different objects or different scenes are the same as those of the aforementioned regions to be stretched corresponding to different temperature ranges.
  • the method and principle of gray-scale stretching are similar or the same, please refer to the foregoing description, and this embodiment will not be repeated here.
  • the target area is the area of interest specified by the user when inputting control instructions or pre-setting the program. It can be any position, shape, size, or size in the image. Any number of image areas.
  • gray-scale stretching is performed on the area to be stretched divided by this method, the area to be stretched can be stretched with a higher tensile strength according to user needs, and other image areas can be stretched with a lower tensile strength. Stretch, or not stretch.
  • Step 202 Determine the grayscale characteristics and the stretching control characteristics of the region to be stretched.
  • step 202 are similar or the same as the method and principle of step 101, please refer to the related record of step 101, which will not be repeated here in this embodiment.
  • Step 203 Determine the dead pixels of the image in the area to be stretched, and correct the gray value of the dead pixels of the image.
  • Image dead pixels refer to pixels whose gray values have abrupt changes compared to the gray values of surrounding pixels, where the abrupt gray values include gray values that are too large or too small than the gray values of surrounding pixels.
  • a sudden change in the gray value may be that the difference between the gray value and the gray value of the surrounding pixels is greater than the preset threshold, and the specific value of the preset threshold can be set as required.
  • the gray value range of the area to be stretched will increase abnormally, thereby affecting the determination of the initial stretching parameters. Therefore, this embodiment of the present embodiment has the image dead pixels in the area to be stretched. Perform correction processing on the gray value of, to improve the accuracy of the initial stretching parameters and improve the image quality after stretching.
  • the gray value of the image's dead pixels can be smoothed in time.
  • the average of the gray values of the pixels around the image's dead pixels is used. The value replaces the gray value of the changed point of the image, or replaces the gray value of the changed point of the image with the median value of the gray value of the pixels around the bad point of the image.
  • Step 204 Determine the initial stretching parameters of the area to be stretched according to the grayscale characteristics of the area to be stretched; adjust the initial stretching parameters according to the stretching control feature to obtain the controlled stretching parameters.
  • step 204 the method and principle of determining the initial stretching parameters of the region to be stretched according to the gray-scale characteristics of the region to be stretched are similar or the same as the method and principle recorded in step 102.
  • the relevant records of step 102 please refer to the relevant records of step 102. The embodiments are not repeated here.
  • this embodiment further includes : Determine part of the area, there are at least two parts of the area, and all the parts constitute the area to be stretched; then according to the mapping relationship between the gray value and the pixel probability density in the part of the area to be stretched, determine the area to be stretched
  • the local stretching parameters include: respectively determining the local stretching parameters corresponding to each partial region according to the mapping relationship between the gray value and the pixel probability density in each partial region of the region to be stretched.
  • the method for determining the partial area can be conventional in the field.
  • the area to be stretched can be divided into multiple partial areas according to the specification of 2 n * 2 n, for example, according to 4*4, 8*8,
  • the 16*16, 32*32 or 64*64 rule divides the area to be stretched into multiple partial areas.
  • the stretch control feature includes at least one of the following: noise, signal-to-noise ratio, or at least part of the edge of the region to be stretched.
  • the initial stretching parameters are adjusted according to the noise to obtain the adjusted stretching parameters: the greater the noise, the smaller the degree of stretching corresponding to the adjusted stretching parameters; when the initial stretching parameters are adjusted according to the signal-to-noise ratio, the adjusted stretching parameters are obtained.
  • stretching parameters the smaller the signal-to-noise ratio, the smaller the degree of stretching corresponding to the adjusted stretching parameters; when the initial stretching parameters are adjusted according to at least part of the edge of the area to be stretched, the adjustment is obtained when the adjusted stretching parameters are obtained
  • the stretching parameter can suppress the stretching degree of at least part of the edge of the stretched area.
  • the stretching degree which can effectively prevent the noise from being excessively stretched and affect the image quality after stretching; the stretching degree of at least part of the edge of the area to be stretched can be suppressed. It effectively suppresses the pot-lid effect of the stretched image, avoids the problem of excessively bright edges of the stretched image, and can significantly improve the quality of the stretched image.
  • adjusting the initial stretching parameters according to the stretching control characteristics to obtain the controlled stretching parameters includes: adjusting according to the stretching Feature, the local stretching parameters corresponding to each partial area are respectively adjusted to obtain the adjustable stretching parameters corresponding to each partial area.
  • the method and principle of adjusting the local stretching parameters corresponding to each partial area are similar or the same as the aforementioned method and principle of adjusting the initial stretching parameters in the global area of the area to be stretched. , I won’t repeat it here.
  • the local stretching parameters corresponding to each partial area can be adjusted to different degrees, so that different partial areas correspond to each other. The different stretching degree that matches with itself further improves the quality of the stretched image.
  • Step 205 Perform a stretching process on the area to be stretched according to the adjusted stretching parameters.
  • performing the stretching process on the area to be stretched according to the adjusted stretching parameter includes: Corresponding to adjust and control the stretching parameters, each part of the area to be stretched is stretched.
  • this embodiment further includes: for the first partial area, according to the adjacent partial area
  • the corresponding control stretch parameter comprehensively adjusts the control stretch parameter corresponding to the first partial area, where the first partial area is any one of the partial areas; according to the control stretch parameters corresponding to each partial area after comprehensive adjustment, Stretching is performed on each part of the area to be stretched.
  • the comprehensive adjustment mentioned here refers to smoothing the control and stretching parameters corresponding to the first partial area according to the control and stretching parameters corresponding to the adjacent partial areas, so that the first partial area corresponds to the control and stretching parameters.
  • the parameters are matched with the adjustment and stretching parameters corresponding to the adjacent partial areas.
  • the execution of this embodiment includes: performing low-frequency area extraction processing on the area to be stretched, where the low-frequency area is that the gray value change rate is less than a preset Threshold region: performing stretching processing on the area to be stretched according to the control and stretching parameters, including: performing stretching processing on the low-frequency area of the area to be stretched according to the control and stretching parameters. Extract the low-frequency area in the area to be stretched, and perform stretching processing on the low-frequency area, so that the high-frequency area or noise in the area to be stretched is not pulled up, which can effectively improve the signal-to-noise ratio of the pulled up image and increase the image Contrast, improve the quality of the image after stretching.
  • the method for extracting the low-frequency region in the region to be stretched may be conventional in the field.
  • the low-frequency region in the region to be stretched can be separated by a wavelet transform method, which will not be repeated here.
  • this embodiment further includes: performing detail extraction processing on the area to be stretched to obtain image details of the area to be stretched; Parameters, after performing the stretching processing on the area to be stretched, the execution includes: adding image details to the stretched image to obtain the processed image.
  • the image details include small targets occupying a few pixels in the area to be stretched or the texture information of the target surface.
  • the image details usually contain the key details of the area to be stretched, but the number of pixels in the image details is small and the gray value changes. The amplitude is small. If the image details are forced to be stretched, the image details may be distorted, and the key details of the region to be stretched will be lost.
  • the region to be stretched is first extracted before the stretch processing is performed.
  • the image details in the region to be stretched are extracted, and the extracted image details are re-added to the stretched image after the stretch processing is performed on the region to be stretched.
  • the method for extracting details of the area to be stretched may be conventional in the field. For example, a bilateral filter may be used to separate the image details in the area to be stretched, which will not be repeated here.
  • this embodiment further includes: performing smoothing processing on the image after the stretching processing.
  • the method of smoothing the stretched image can be conventional in the field.
  • a bilateral filter can be used to smooth the edge of the partially stretched image to reduce noise, so as to reduce the local stretched image.
  • the edge of the image is halo to achieve the effect of maintaining the edges and reducing the noise;
  • the dither dithering algorithm can be used to smooth the stretched image to avoid step breaks between the stretched image areas.
  • the method of this embodiment includes different stretching modes, and the different stretching modes correspond to different stretching control characteristics.
  • the stretching mode includes a high-gain mode and a low-gain mode, or a low-contrast mode and a high-contrast mode.
  • the stretched image is sent to the display device for the display device to display the stretched image
  • the display device may include setting The display screen, display, etc. in the control terminal.
  • the operator can perform different applications according to the stretched image displayed in the display device, for example, target tracking, infrared ranging, infrared temperature measurement, remote control, and real-time based on the stretched image displayed in the display device Surveillance, etc., for example, the infrared camera set in the drone takes the infrared image of the base station, and the image processing module set in the drone performs the gray-scale stretch processing on the infrared image, and the stretched The image is sent to the display device of the control terminal, and the operator judges the temperature of the base station according to the image displayed in the display device, thereby realizing the infrared remote temperature measurement of the base station.
  • grayscale stretching the image contrast is enhanced, the image details are obvious, and the image quality can be effectively improved. According to the image
  • the grayscale characteristics and stretching control characteristics of the area to be stretched by determining the area to be stretched; determining the gray-scale characteristics and stretching control characteristics of the area to be stretched; determining the initial stretching parameters of the area to be stretched according to the gray-scale characteristics of the area to be stretched; Stretch control characteristics, adjust the initial stretch parameters to obtain the regulated stretch parameters; according to the regulated stretch parameters, stretch the area to be stretched.
  • the grayscale characteristics of the image dead pixels in the area to be stretched are corrected, which is beneficial to improve the initial stretching parameters and the area to be stretched.
  • the matching degree of the region according to the characteristics of the stretching control, the initial stretching parameters are adjusted in various forms, and the adjusted stretching parameters obtained have the advantages of strong applicability, high flexibility, and perfect matching.
  • the area to be stretched is adapted
  • Fig. 2a is another image processing method provided by an embodiment of the present application, as shown in Fig. 2a, including:
  • S1 Receive the input image, detect the gray value range of the input image, and correct the gray value of the image defect in the input image; determine the first initial stretching parameter according to the detected gray value range; The intensity modulates the first initial stretching parameter to obtain the first modulated stretch parameter; according to the first modulated stretch parameter, gray-scale stretch processing is performed on the input image, specifically, linear stretch processing is performed on the input image.
  • the linear stretching process is to stretch the gray values of all pixels of the input image with the same stretching strength and the same offset amplitude.
  • the linear stretching process is simple to implement, and the stretching efficiency is high, which is conducive to realization. Fast image gray value optimization processing.
  • step S2 Obtain the global area histogram of the input image processed in step S1, and determine the gray value probability density function of the global area according to the global area histogram; determine the gray value of the global area according to the gray value probability density function of the global area Cumulative distribution function; determine the second initial stretching parameter according to the cumulative distribution function of the gray value of the global area; adjust the second initial stretching parameter according to the noise intensity to obtain the second adjustable stretching parameter; according to the second adjustable stretching Parameter, the input image processed in step S1 is subjected to global area adaptive stretching processing. According to the determined cumulative distribution function of the gray value of the global area, the second initial stretching parameter is determined, and the second initial stretching parameter is adaptable to the gray value of the pixel in the global area.
  • the global area of the input image is subjected to gray value stretching processing, which realizes the adaptive stretching processing on the gray value of different pixels, and the gray scale stretching effect is obvious , which is beneficial to effectively increase the contrast of infrared images and improve the image quality of infrared images.
  • step S3 Divide the input image processed in step S2 into multiple partial regions, and obtain a partial region histogram of each partial region respectively; determine the gray value probability density function of each partial region according to each partial region histogram ; According to the gray value probability density function of each partial area, determine the gray value cumulative distribution function of each partial area; according to the gray value cumulative distribution function of each partial area, determine the third initial pull of each partial area Stretching parameters; according to the noise intensity of each partial area, the third initial stretching parameter of each partial area is adjusted to obtain the third adjustable stretching parameter of each partial area; according to the third adjustable stretching parameter of each partial area Stretching parameters, adaptive stretch processing for each partial area respectively.
  • the third initial stretching parameter of each partial area is determined, and the third initial stretching parameter of each partial area is compatible with the pixel gray value of the corresponding partial area. sex.
  • each partial area is separately subjected to gray-scale stretching processing to realize different pixel points of different partial areas
  • the gray value of the infrared image is adaptively stretched, and the variety of stretching processing methods is rich, the gray-scale stretching effect is significant, the contrast of the infrared image can be effectively improved, and the image quality of the infrared image can also be effectively improved.
  • gray-scale stretch processing methods of S1, S2, and S3 can be used at the same time, or only one or two gray-scale stretch processing methods can be used.
  • the order of using the degree of stretching processing method can be adjusted arbitrarily.
  • different gray-scale stretch processing methods can be used for different sub-regions of the area to be stretched.
  • the area to be stretched can be divided into different temperature ranges. Different sub-regions perform different gray scale stretching processing on different sub-regions.
  • the gray-scale stretching processing methods for infrared images are diverse and flexible. When the infrared image is enhanced, any combination of stretching processing methods can be carried out according to application requirements or user needs, which is conducive to realizing effective infrared images. Contrast enhancement processing helps to significantly improve the image quality of infrared images.
  • FIG. 3 is a schematic structural diagram of an image processing apparatus provided by an embodiment of the application. As shown in FIG. 3, an embodiment of the present application provides an image processing apparatus, which specifically includes a processor 301 and a memory 302, wherein,
  • the memory 302 is configured to store instructions that can be executed by the processor.
  • the processor 301 is configured to execute instructions stored in the memory 302 to implement the image processing method in the embodiment shown in FIG. 1 to FIG. 2.
  • processor 301 executes instructions to perform the following steps:
  • the initial stretching parameters are adjusted to obtain the adjusted stretching parameters
  • the stretched area is stretched.
  • the processor 301 determines the initial stretching parameters of the area to be stretched according to the grayscale characteristics of the area to be stretched, including at least one of the following:
  • the local stretch parameter of the area to be stretched is determined.
  • the local stretch parameter constitutes the initial stretch parameter, where the local stretch parameter and Corresponding to some areas.
  • the execution includes:
  • the stretching control feature includes at least one of the following: noise, signal-to-noise ratio, or at least part of the edge of the region to be stretched.
  • the processor 301 adjusts the initial stretching parameters according to the noise to obtain the adjusted stretching parameters: the greater the noise, the smaller the degree of stretching corresponding to the adjusted stretching parameters;
  • the processor 301 adjusts the initial stretching parameters according to the signal-to-noise ratio to obtain the adjusted stretching parameters: the smaller the signal-to-noise ratio, the smaller the degree of stretching corresponding to the adjusted stretching parameters;
  • the processor 301 adjusts the initial stretching parameters according to at least part of the edges of the area to be stretched to obtain the adjusted stretching parameters
  • the obtained adjusted stretching parameters can suppress the stretching degree of at least part of the edges of the area to be stretched.
  • the execution includes: determining a partial region, and the partial region is at least two regions. One, all part of the area constitutes the area to be stretched;
  • the processor 301 determines the local stretching parameters of the area to be stretched according to the mapping relationship between the gray value in the partial area of the area to be stretched and the pixel probability density, including: The mapping relationship between the degree value and the pixel probability density determines the local stretching parameters corresponding to each partial area.
  • the processor 301 adjusts the initial stretching parameters according to the stretching control characteristics to obtain the controlled stretching parameters, including: according to the stretching control characteristics, respectively adjust the local stretching parameters corresponding to each partial area to obtain each partial area Corresponding control and stretching parameters;
  • the processor 301 performs stretching processing on the area to be stretched according to the adjusted stretching parameter, including: performing stretching processing on each partial area of the area to be stretched according to the adjusted stretching parameter corresponding to each partial area.
  • the processor 301 performs stretching processing on each partial area of the to-be-stretched area according to the adjusted stretching parameters corresponding to each partial area, and further includes:
  • the first partial area comprehensively adjust the control and stretching parameters corresponding to the first partial area according to the control and stretching parameters corresponding to the adjacent partial areas, where the first partial area is any one of the partial areas;
  • each partial area of the to-be-stretched area is stretched.
  • the processor 301 Before determining the grayscale characteristics of the area to be stretched, the processor 301 further executes the following steps: determining the area to be stretched.
  • the area to be stretched can be determined according to the input control instruction, or according to preset The procedure is OK.
  • the input control command includes at least one of the following:
  • the processor 301 performs the following steps before performing the stretching processing on the area to be stretched according to the adjusted stretching parameters:
  • the processor 301 performs stretching processing on the area to be stretched according to the adjusted stretching parameters, including:
  • the low-frequency region of the region to be stretched is stretched.
  • the processing further includes:
  • the execution includes:
  • the processor 301 After the processor 301 performs the stretching processing on the area to be stretched, the processor 301 further executes the following steps: smoothing the image after the stretching processing.
  • the stretching process includes different stretching modes, and different stretching modes correspond to different stretching control characteristics.
  • the execution includes: sending the stretched image to the display device, and the display device is used to display the stretched image.
  • the image processing apparatus may further include a communication interface 303 and a bus 304, where the processor 301, the memory 302, and the communication interface 303 may be connected to each other through the bus 304.
  • the communication interface 303 may include an interface for communicating between an image processing device and an infrared camera, or with a control terminal.
  • the bus 304 may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc.
  • PCI Peripheral Component Interconnect
  • EISA Extended Industry Standard Architecture
  • the above-mentioned bus 304 can be divided into an address bus, a data bus, a control bus, and the like. For ease of presentation, only a thick line is used in FIG. 3 to represent it, but it does not mean that there is only one bus or one type of bus.
  • the above-mentioned embodiments can refer to each other and learn from each other, and the same or similar steps and nouns will not be repeated one by one.
  • part or all of the above modules can also be implemented by embedding on a certain chip of the image processing device in the form of an integrated circuit. And they can be implemented separately or integrated together. That is, the above modules can be configured as one or more integrated circuits that implement the above methods, for example: one or more specific integrated circuits (Application Specific Integrated Circuit, ASIC for short), or one or more microprocessors (Digital Singnal Processor, DSP for short), or, one or more Field Programmable Gate Arrays (FPGA for short).
  • ASIC Application Specific Integrated Circuit
  • DSP Digital Singnal Processor
  • FPGA Field Programmable Gate Arrays
  • the grayscale characteristics and stretching control characteristics of the area to be stretched by determining the area to be stretched; determining the gray-scale characteristics and stretching control characteristics of the area to be stretched; determining the initial stretching parameters of the area to be stretched according to the gray-scale characteristics of the area to be stretched; Stretch control characteristics, adjust the initial stretch parameters to obtain the regulated stretch parameters; according to the regulated stretch parameters, stretch the area to be stretched.
  • the grayscale characteristics of the image dead pixels in the area to be stretched are corrected, which is beneficial to improve the initial stretching parameters and the area to be stretched.
  • the matching degree of the region according to the characteristics of the stretching control, the initial stretching parameters are adjusted in various forms, and the adjusted stretching parameters obtained have the advantages of strong applicability, high flexibility, and perfect matching.
  • the area to be stretched is adapted
  • FIG. 4 is a schematic structural diagram of an image processing system provided by an embodiment of the application.
  • the image processing system includes an infrared camera and the image processing device shown in FIG. 3, wherein the infrared camera and the image processing device Corresponding to the connection, the image processing device is used to process the infrared image collected by the infrared camera.
  • the infrared camera is mounted on a movable platform, and the movable platform is used to perform infrared image collection tasks; the infrared camera is used to collect infrared images when the movable platform performs infrared image collection tasks, and send the infrared images to the image Processing device:
  • the image processing device is used to receive infrared images taken by an infrared camera and process the infrared images.
  • Movable platforms include unmanned vehicles, drones, and mobile robots.
  • the image processing system further includes a display device for displaying the image processed by the image processing device.
  • the image processing further includes a control terminal, and the control terminal may send a control instruction to the image processing apparatus, and the control instruction is used to determine the area to be stretched or determine the image stretch mode.
  • the display device is provided on the control terminal.
  • the above embodiments it may be implemented in whole or in part by software, hardware, firmware, or any combination thereof.
  • software it can be implemented in the form of a computer program product in whole or in part.
  • the computer program product includes one or more computer instructions.
  • the computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.
  • the computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium.
  • coaxial cable, optical fiber, digital subscriber line (digital subscriber line, DSL)) or wireless (for example, infrared, wireless, microwave, etc.) means to transmit to another website, computer, image processing equipment, or data center.
  • the computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as an image processing device or a data center integrated with one or more available media.
  • the usable medium may be a magnetic medium (for example, a floppy disk, a hard disk, and a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, a solid state disk (SSD)).
  • Computer-readable media include computer storage media and communication media, where communication media includes any media that facilitates the transfer of computer programs from one place to another.
  • the storage medium may be any available medium that can be accessed by a general-purpose or special-purpose computer.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Image Processing (AREA)

Abstract

一种图像处理方法、装置、系统和存储介质，确定待拉伸区域的灰度特征和拉伸调控特征，待拉伸区域对应输入图像的至少部分区域(101)；根据所述待拉伸区域的灰度特征，确定所述待拉伸区域的初始拉伸参数(102)；根据所述拉伸调控特征，对所述初始拉伸参数进行调控，得到调控拉伸参数(103)；根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理(104)。在根据待拉伸区域的灰度特征确定出初始拉伸参数后，根据拉伸调控特征对初始拉伸参数进行调控，得到调控拉伸参数，然后根据调控拉伸参数对待拉伸区域进行灰度拉伸处理，调控拉伸参数与待拉伸区域的匹配度高，适用性强，处理后的红外图像的图像质量高，图像增强效果显著。

Description

图像处理方法、装置、系统和存储介质 技术领域
本申请涉及图像处理技术领域，尤其涉及一种图像处理方法、装置、系统和存储介质。
背景技术
红外图像是通过获取物体的红外光强度而形成的图像，其存在对比度低、灰度直方图集中、噪声跳动明显等一些影响图像质量的问题，因此常需要对红外图像进行处理。
相关地，在对红外图像进行处理时，根据预先确定的灰度拉伸参数，对红外图像进行的对比度拉伸处理。
然而，根据预先确定的灰度拉伸参数，对红外图像进行的对比度拉伸处理，存在灰度拉伸参数可能无法适用于红外图像的某些局部区域的问题，这导致红外图像的处理效果不佳、图像质量受损的问题。
发明内容
本申请提供一种图像处理方法、装置、系统和存储介质，用以解决现有的图像处理技术存在的红外图像处理效果不佳、图像质量受损的问题。
第一方面，本申请提供一种图像处理方法，所述方法用于红外图像的处理，所述方法包括：
确定待拉伸区域的灰度特征和拉伸调控特征，所述待拉伸区域对应输入图像的至少部分区域；
根据所述待拉伸区域的灰度特征，确定所述待拉伸区域的初始拉伸参数；
根据所述拉伸调控特征，对所述初始拉伸参数进行调控，得到调控拉伸参数；
根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理。
第二方面，本申请提供一种图像处理装置，所述装置用于红外图像的处理，所述装置包括处理器和存储器，所述处理器与所述存储器通信连接，所述存储器存储有可被所述处理器执行的指令，所述处理器执行所述指令 以执行以下步骤：
确定待拉伸区域的灰度特征和拉伸调控特征，所述待拉伸区域对应输入图像的至少部分区域；
根据所述待拉伸区域的灰度特征，确定所述待拉伸区域的初始拉伸参数；
根据所述拉伸调控特征，对所述初始拉伸参数进行调控，得到调控拉伸参数；
根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理。
第三方面，本申请提供一种图像处理系统，包括红外相机和如第二方面所述的图像处理装置，其中，所述红外相机与所述图像处理装置对应连接，所述图像处理装置用于对所述红外相机采集的红外图像进行处理。
第四方面，本申请提供一种计算机可读存储介质，其上存储有计算机程序，计算机程序被处理器执行以实现如上任一项的方法。
本申请提供一种图像处理方法、装置、设备和存储介质，通过确定待拉伸区域的灰度特征和拉伸调控特征，所述待拉伸区域对应输入图像的至少部分区域；根据所述待拉伸区域的灰度特征，确定所述待拉伸区域的初始拉伸参数；根据所述拉伸调控特征，对所述初始拉伸参数进行调控，得到调控拉伸参数；根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理。在根据待拉伸区域的灰度特征确定出初始拉伸参数后，根据拉伸调控特征对初始拉伸参数进行调控，得到调控拉伸参数，然后根据调控拉伸参数对待拉伸区域进行灰度拉伸处理，调控拉伸参数与待拉伸区域的匹配度高，适用性强，处理后的红外图像的图像质量高，图像增强效果显著。
附图说明
此处的附图被并入说明书中并构成本说明书的一部分，示出了符合本公开的实施例，并与说明书一起用于解释本公开的原理，但并不对本申请构成限制。
图1为本申请实施例提供的一种图像处理方法的流程示意图；
图1a为本申请实施例提供的一种根据待拉伸区域的灰度值范围确定初始拉伸参数的示意图；
图1b为本申请实施例提供的一种根据待拉伸区域的灰度直方图确定灰 度概率密度函数的示意图；
图2为本申请实施例提供的又一种图像处理方法的流程示意图；
图2a为本申请实施例提供的再一种图像处理方法的流程示意图；
图3为本申请实施例提供的一种图像处理装置的结构示意图；
图4为本申请实施例提供的一种图像处理系统的结构示意图。
通过上述附图，已示出本公开明确的实施例，后文中将有更详细的描述。这些附图和文字描述并不是为了通过任何方式限制本公开构思的范围，而是通过参考特定实施例为本领域技术人员说明本公开的概念。
具体实施方式
这里将详细地对示例性实施例进行说明，其示例表示在附图中。下面的描述涉及附图时，除非另有表示，不同附图中的相同数字表示相同或相似的要素。以下示例性实施例中所描述的实施方式并不代表与本公开相一致的所有实施方式。相反，它们仅是与如所附权利要求书中所详述的、本公开的一些方面相一致的装置和方法的例子。
本申请涉及的名词解释：
信噪比：Signal-Noise Ratio，简称SNR或S/N,是信号均值与背景信号值方差的比值,在图像中，信号均值常指灰度值，背景信号值方差的物理意义是噪声功率。
本申请具体的应用场景为：红外图像是通过获取物体的红外光强度而形成的图像，在红外成像领域，红外传感器的直出图像存在对比度低、灰度直方图集中、噪声跳动明显等一些影响图像质量的问题。示例性地，由设置在无人机上的红外相机采集到的红外图像，存在信噪比低、温度展示不充分、存在暗处过暗、亮处过曝的问题，其无法展示出场景物体的细节，在需要根据采集的红外图像进行实时监控、目标追踪、远程遥控、红外测距、红外测温等应用时，红外图像的图像精度无法满足应用要求，因此本申请实施例对红外图像进行了增强处理，以获得符合要求的红外图像。
上述应用场景中，现有技术常采用的红外图像增强处理方法包括：根据预先确定的灰度拉伸参数，对红外图像进行全局区域的对比度拉伸处理。然而，根据预先确定的灰度拉伸参数，对红外图像进行全局区域的对比度拉伸处理，存在灰度拉伸参数可能无法适用于红外图像的部分局部区域的 问题，例如，预先确定的灰度拉伸参数不适用于红外图像的边角区域，导致灰度拉伸处理后的红外图像的边角区域更亮，或者，预先确定的灰度拉伸参数不适用于红外图像的噪声区域，对噪声区域和正常区域采用同一灰度拉伸参数进行灰度拉伸处理，导致噪声区域灰度值更高，图像闪烁更明显，这些都导致红外图像的增强处理效果不佳、图像质量受损的问题。
本申请提出一种图像处理方法、装置、系统和存储介质，旨在解决上述技术问题。
图1为本申请实施例提供的一种图像处理方法的流程示意图，本方法用于红外图像的处理，如图1所示，该方法包括：
步骤101、确定待拉伸区域的灰度特征和拉伸调控特征，待拉伸区域对应输入图像的至少部分区域。
在本实施例中，具体地，本实施例的执行主体为图像处理模块，图像处理模块设置于控制终端或可移动平台中，可移动平台包括无人机、无人车、可移动机器人等。本实施例以执行主体为设置在无人机中的图像处理模块为例进行说明。
待拉伸区域是输入图像中需要进行灰度拉伸的区域，其对应输入图像的至少部分区域，具体可以是输入图像的全部区域，也可以是输入图像的部分区域。例如，待拉伸区域可以是一整张输入图像，也可以是输入图像的部分区域，或者是输入图像中某一物体对应的区域，或者是某一场景对应的区域，或者是某一温度区间对应的区域，也可以是输入图像中用户指定的感兴趣区域，也可以是输入图像中两个场景的交界区域。灰度特征能够反映图像的像素点的灰度分布情况，具体可以包括灰度值、灰度值范围、灰度直方图、灰度概率密度函数等特征。图像的像素点拥有的最大灰度级范围为[0,255]，灰度值为0时，其对应的颜色为白色；灰度值为255时，其对应的颜色为黑色。
拉伸调控特征表征用于调控拉伸参数的特征，包括但不限于输入图像的噪声、信噪比、或待拉伸区域的至少部分边缘。由红外相机或红外采集器采集的红外图像通常包括噪声，噪声可以包括单点随机噪声、横条纹噪声、竖条纹噪声、热背景噪声中的任意一种或多种，噪声影响红外图像的 图像质量，容易造成图像伪影。输入图像的信噪比为像素灰度值与噪声功率的比值。待拉伸区域的至少部分边缘可以通过边缘检测算子检测得到，边缘检测算子包括Roberts Cross算子、Prewitt算子、Sobel算子、Canny算子、Marr-Hildreth算子等。
步骤102、根据待拉伸区域的灰度特征，确定待拉伸区域的初始拉伸参数。
在本实施例中，具体地，拉伸参数可以包括拉伸强度、灰度偏移幅度等，根据待拉伸区域的灰度特征，确定待拉伸区域的初始拉伸参数，包括以下至少一种：根据待拉伸区域的灰度值范围，确定待拉伸区域的初始拉伸参数；根据待拉伸区域的全局区域中灰度值与像素概率密度之间的映射关系，确定待拉伸区域的初始拉伸参数；根据待拉伸区域的部分区域中灰度值与像素概率密度之间的映射关系，确定待拉伸区域的局部拉伸参数，局部拉伸参数构成初始拉伸参数，其中，局部拉伸参数与部分区域相对应。
根据待拉伸区域的灰度值范围，确定待拉伸区域的初始拉伸参数，示例性地，图1a是本实施例提供的一种根据待拉伸区域的灰度值范围确定初始拉伸参数的示意图，如图1a所示，f(x，y)表示待拉伸区域中某一像素点(x,y)的灰度值，g(x，y)表示待拉伸区域中同一像素点(x,y)被拉伸后的灰度值，待拉伸区域的灰度值范围为[a，b]，期望拉伸得到的灰度值范围为[m，n]，其中(n-m)＞(b-a)，拉伸过程可以表示如下：
Figure PCTCN2020087574-appb-000001
式(1)中，(n-m)/(b-a)即为根据待拉伸区域的灰度值范围确定的初始拉伸参数。
根据待拉伸区域的全局区域中灰度值与像素概率密度之间的映射关系，确定待拉伸区域的初始拉伸参数。其中，灰度值与像素概率密度之间的映射关系可以是灰度值的像素概率密度函数，也可以是灰度值的累积分布函数。优选地，灰度值与像素概率密度之间的映射关系为灰度值的累积分布函数。示例性地，本实施例中可以先获取待拉伸区域的全局区域的灰度直方图，再根据灰度直方图确定灰度值的概率密度函数，然后根据灰度值的概率密度函数确定灰度值的累积分布函数，最后根据灰度值的累积分布函数确定待拉伸区域的初始拉伸参数。图1b是本实施例提供的一种根 据待拉伸区域的灰度直方图确定灰度概率密度函数的示意图,如图1b所示，对于待拉伸区域来说，其灰度统计直方图是一个离散函数，设待拉伸区域总的像素数为N，具有的灰度值的级数为L级，则直方图可以表示为：
h(r i)＝n i      式(2)
式(2)中，i在灰度值级数范围[0，L-1]内取值，r i表示第i个灰度值级数对应的灰度值，n i表示灰度值为r i的像素个数。根据式(2)表示的灰度直方图确定灰度值的概率密度函数为：
Figure PCTCN2020087574-appb-000002
式(3)中，p(r i)表示灰度值r i出现的概率。对式(3)所示的灰度值的概率密度函数进行积分即可得到灰度值的累积分布函数，进而根据灰度值的累积分布函数确定待拉伸区域的初始拉伸参数。
根据待拉伸区域的部分区域中灰度值与像素概率密度之间的映射关系，确定待拉伸区域的局部拉伸参数，局部拉伸参数构成初始拉伸参数，其中，局部拉伸参数与部分区域相对应。其中，上述部分区域是将待拉伸区域的全局区域进行分块后得到的，例如，可以将待拉伸区域的全局区域按照4*4、8*8、16*16、32*32或者64*64等规格进行分块，得到多块上述部分区域。在每一块部分区域中根据灰度值与像素概率密度之间的映射关系，确定与待拉伸区域中的每一块部分区域分别对应的多个局部拉伸参数，多个局部拉伸参数构成待拉伸区域的初始拉伸参数。在每一块部分区域中根据灰度值与像素概率密度之间的映射关系确定与该部分区域对应的局部拉伸参数的方法和原理，与前述的根据待拉伸区域的全局区域中灰度值与像素概率密度之间的映射关系确定待拉伸区域的初始拉伸参数的方法和原理相似或相同，参加前述记载，本实施例不再赘述。
上述三种方法仅用于对本实施例的解释说明，并不用于限制本申请，本申请还可采用其它的方法来确定待拉伸区域的初始拉伸参数，本实施例在此不再赘述。
步骤103、根据拉伸调控特征，对初始拉伸参数进行调控，得到调控拉伸参数。
在本实施例中，具体地，拉伸调控特征包括输入图像的噪声、信噪比、或待拉伸区域的至少部分边缘。步骤102得到的初始拉伸参数，仅考虑了 待拉伸区域的灰度特征对拉伸参数的影响，但是在实际的拉伸过程中，为了确保拉伸后能够获得更好的图像质量，在确定拉伸参数时还需要考虑其它因素的影响，例如，待拉伸区域红外图像的噪声、信噪比、或待拉伸区域的至少部分边缘等。示例性地，若拉伸强度过大，噪声也会被过度拉伸，从而影响拉伸后的图像质量；或者，红外图像的边角会随着时间的推移而发亮，形成中心较暗、边角较亮的锅盖效应，此时若对整张图像采用同样的拉伸参数进行拉伸，拉伸得到的红外图像的中心部分和边角部分的图像质量将存在较大差异。因此，本实施例中利用拉伸调控特征对步骤102得到的初始拉伸参数进行调控，以得到拉伸效果更好的拉伸参数。
根据拉伸调控特征，对初始拉伸参数进行调控，示例性地，可以根据拉伸前的噪声强度对初始拉伸参数进行调控，以使拉伸后的噪声强度较小，或者拉伸后的红外图像的信噪比较大，从而提升拉伸后的红外图像的图像高质量；或者，可以根据待拉伸区域的至少部分边缘的灰度值，对初始拉伸参数进行调控，以使拉伸后的红外图像的中心部分和边角部分的图像质量相似或相同。
步骤104、根据调控拉伸参数，对待拉伸区域进行拉伸处理。
在本实施例中，具体地，根据调控拉伸参数，对待拉伸区域的灰度值进行拉伸处理，根据调控拉伸参数对待拉伸区域进行灰度值拉伸处理的方法，可以采用本领域的常规灰度值拉伸方法，例如，可以采用领域内抑制的算法程序对待拉伸区域进行灰度值拉伸处理，本实施例在此不再赘述。
本实施例中，通过确定待拉伸区域的灰度特征和拉伸调控特征，待拉伸区域对应输入图像的至少部分区域；根据待拉伸区域的灰度特征，确定待拉伸区域的初始拉伸参数；根据拉伸调控特征，对初始拉伸参数进行调控，得到调控拉伸参数；根据调控拉伸参数，对待拉伸区域进行拉伸处理。在根据待拉伸区域的灰度特征确定出初始拉伸参数后，根据拉伸调控特征对初始拉伸参数进行调控，得到调控拉伸参数，然后根据调控拉伸参数对待拉伸区域进行灰度拉伸处理，调控拉伸参数与待拉伸区域的匹配度高，适用性强，处理后的红外图像的图像质量高，图像增强效果显著。
图2为本申请实施例提供的又一种图像处理方法的流程示意图，在图1 的基础上，如图2所示，该方法包括：
步骤201、确定待拉伸区域。
在本实施例中，具体地，待拉伸区域是输入图像中需要进行灰度拉伸处理的区域，其对应输入图像的部分至少区域，待拉伸区域可以是输入图像的全局区域，或者是输入图像被分割后得到的部分区域，或者是输入图像中某一物体对应的区域，或者是某一场景对应的区域，或者是某一温度区间对应的区域，或者是用户指定的感兴趣区域等。待拉伸区域的数量至少为一个。
确定待拉伸区域的方法可以有多种，例如，可以通过机器学习方法获取待拉伸区域，具体的，当需要对红外图像中某一物体(例如人或其他景物)的图像效果进行增强时，可以通过机器学习算法对输入图像进行识别，识别出需要增强的物体所在的区域，将其作为待拉伸区域。也可以根据输入的控制指令确定待拉伸区域，或者根据预先设置的程序确定待拉伸区域。其中，控制指令可以是在确定待拉伸区域之前由用户输入的，预先设置的程序可以是预置在存储器里的，也可以是在确定待拉伸区域之前由用户预先设置的。输入的控制指令或者预先设置的程序用于指示确定待拉伸区域的规则或者要求，示例性地，输入的控制指令包括以下至少一种：根据图像温度划分待拉伸区域；根据图像中的物体划分待拉伸区域；根据图像中的场景划分待拉伸区域；将输入图像中指定的部分区域划分为待拉伸区域，该部分区域可以是用户的感兴趣区域，可以是用户输入的控制指令所指示的任意区域，例如可以是图像的上半部分区域或者右下角四分之一区域等。
根据图像温度划分待拉伸区域时，可以将某一目标温度区间对应的图像区域作为待拉伸区域，也可以将不同温度区间对应的不同图像区域作为多个待拉伸区域。当将某一目标温度区间对应的图像区域作为待拉伸区域时，待拉伸区域可以根据用户需求采用较高的拉伸强度进行拉伸，其它温度区间对应的图像区域采用较低的拉伸强度进行拉伸，或者不拉伸，从而提升目标温度区间对应的拉伸后图像的对比度。当根据不同的温度区间将输入图像划分为多个不同的待拉伸区域时，可以对不同的温度区间对应的待拉伸区域采用不同的拉伸参数进行拉伸，以提升拉伸后图像的对比度，例如，对于用户不感兴趣的超低温区间对应的待拉伸区域，可以采用较低 的拉伸强度进行拉伸；对于低温区间或者中等温度区间对应的待拉伸区域，可以较超低温区间适当提升拉伸强度后进行拉伸；对于高温区间对应的待拉伸区域，可以根据不同的场景和/或用户需求，采用较高的拉伸强度进行拉伸。
同理，根据图像中的物体划分待拉伸区域时，可以仅将用户感兴趣的物体对应的图像区域划分为待拉伸区域，也可以将图像中不同的物体分别对应的图像区域划分为不同的待拉伸区域；根据图像中的场景划分待拉伸区域时，可以仅将用户感兴趣的场景对应的图像区域划分为待拉伸区域，也可以将图像中不同的场景分别对应的图像区域划分为不同的待拉伸区域。其中，在上述两种待拉伸区域的划分方法中，对不同物体或不同场景对应的待拉伸区域进行灰度拉伸的方法和原理，与前述的对不同温度区间对应的待拉伸区域进行灰度拉伸的方法和原理相似或相同，参见前述记载，本实施例在此不再赘述。
直接将指定的图像中的目标区域划分为待拉伸区域时，该目标区域是用户在输入控制指令或者预先设置程序时指定的感兴趣区域，可以是图像中任意位置、任意形状、任意大小、任意数量的图像区域。在对本方法划分得到的待拉伸区域进行灰度拉伸时，待拉伸区域可以根据用户需求而采用较高的拉伸强度进行拉伸，其它图像区域可以采用较低的拉伸强度进行拉伸，或者不拉伸。
步骤202、确定待拉伸区域的灰度特征和拉伸调控特征。
步骤202的方法和原理，与步骤101的方法和原理相似或相同，参见步骤101的相关记载，本实施例在此不再赘述。
步骤203、确定待拉伸区域中的图像坏点，并对图像坏点的灰度值进行校正处理。
图像坏点是指灰度值较周围像素点的灰度值发生突变的像素点，其中，灰度值发生突变包括灰度值较周围像素点的灰度值过大或者过小，具体地，灰度值发生突变可以是灰度值与周围像素点的灰度值的差值大于预设阈值，预设阈值的具体取值可以根据需要进行设定。当待拉伸区域中存在图像坏点时，会导致待拉伸区域的灰度值范围异常增大，进而影响初始拉伸参数的确定，因此，本实施例对待拉伸区域中的图像坏点的灰度值进行校正处理，以 提升初始拉伸参数的准确性，提升拉伸后的图像质量。
可以采用多种方法对图像坏点的灰度值进行校正处理，例如，可以对图像坏点的灰度值进行时域平滑处理，优选地，利用图像坏点周围像素点的灰度值的平均值替换图像换点的灰度值，或者利用图像坏点周围像素点的灰度值的中位值替换图像换点的灰度值。
上述对图像坏点的灰度值进行校正处理的方法仅用于对本实施例进行解释说明，并不用于限制本申请，本申请还可采用其它方法对图像坏点的灰度值进行校正处理，此处不再一一赘述。
步骤204、根据待拉伸区域的灰度特征，确定待拉伸区域的初始拉伸参数；根据拉伸调控特征，对初始拉伸参数进行调控，得到调控拉伸参数。
步骤204中，根据待拉伸区域的灰度特征确定待拉伸区域的初始拉伸参数的方法和原理，与步骤102中记载的方法和原理相似或相同，详见步骤102的相关记载，本实施例在此不再赘述。
在步骤102的基础上，可选地，根据待拉伸区域的部分区域中灰度值与像素概率密度之间的映射关系，确定待拉伸区域的局部拉伸参数之前，本实施例还包括：确定部分区域，部分区域至少为两个，全部部分区域构成待拉伸区域；则根据待拉伸区域的部分区域中灰度值与像素概率密度之间的映射关系，确定待拉伸区域的局部拉伸参数，包括：分别根据待拉伸区域的各个部分区域中灰度值与像素概率密度之间的映射关系，确定各个部分区域所对应的局部拉伸参数。其中，确定部分区域的方法可以是领域内常规的，例如，可以按照2 n*2 n的规格将待拉伸区域划分成多个部分区域，示例性地，按照4*4、8*8、16*16、32*32或者64*64的规则将待拉伸区域等分成多个部分区域。
可选地，本实施例中，拉伸调控特征包括以下至少一种：噪声、信噪比、或待拉伸区域的至少部分边缘。当根据噪声对初始拉伸参数进行调控，得到调控拉伸参数时：噪声越大，调控拉伸参数对应的拉伸程度越小；当根据信噪比对初始拉伸参数进行调控，得到调控拉伸参数时：信噪比越小，调控拉伸参数对应的拉伸程度越小；当根据待拉伸区域的至少部分边缘对初始拉伸参数进行调控，得到调控拉伸参数时，得到的调控拉伸参数能够对待拉伸区域的至少部分边缘的拉伸程度进行抑制。当噪声过大或者信噪 比过小时，适当抑制拉伸程度，可以有效避免噪声被过度拉伸而影响拉伸后的图像质量；对待拉伸区域的至少部分边缘的拉伸程度进行抑制，能够有效抑制拉伸后图像的锅盖效应，避免拉伸后图像边缘过亮的问题出现，能够显著提升拉升后的图像质量。
可选地，本实施例中，当初始拉伸参数由前述的局部拉伸参数构成时，根据拉伸调控特征，对初始拉伸参数进行调控，得到调控拉伸参数，包括：根据拉伸调控特征，分别对各个部分区域所对应的局部拉伸参数进行调控，得到各个部分区域所对应的调控拉伸参数。在各个部分区域中，分别对各个部分区域对应的局部拉伸参数进行调控的方法和原理，与前述的在待拉伸区域的全局区域中进行初始拉伸参数进行调控的方法和原理相似或相同，此处不再赘述。分别对各个部分区域所对应的局部拉伸参数进行调控，可以根据各个部分区域的实际情况，根据用户需求，对各个部分区域对应的局部拉伸参数进行不同程度的调控，使得不同的部分区域对应不同的、与自身匹配度较高的拉伸程度，进一步提升拉伸后图像的质量。
步骤205、根据调控拉伸参数，对待拉伸区域进行拉伸处理。
可选地，当待拉伸区域包括多个部分区域，且各个部分区域对应不同的调控拉伸参数时，根据调控拉伸参数，对待拉伸区域进行拉伸处理，包括：根据各个部分区域所对应的调控拉伸参数，对待拉伸区域的各个部分区域进行拉伸处理。
可选地，在根据各个部分区域所对应的调控拉伸参数，对待拉伸区域的各个部分区域进行拉伸处理之前，本实施例还包括：对于第一部分区域，根据与其相邻的部分区域所对应的调控拉伸参数对第一部分区域所对应的调控拉伸参数进行综合调整，其中，第一部分区域为部分区域中的任意一个；根据综合调整后的各个部分区域所对应的调控拉伸参数，对待拉伸区域的各个部分区域进行拉伸处理。其中，这里所说的综合调整，是指根据相邻的部分区域所对应的调控拉伸参数，对第一部分区域所对应的调控拉伸参数进行平滑处理，以使第一部分区域对应的调控拉伸参数与相邻的部分区域对应的调控拉伸参数相匹配。通过前述的综合调整，可以避免拉伸后的相邻部分区域之间产生明显的灰度值跳变现象。
可选地，在根据调控拉伸参数，对待拉伸区域进行拉伸处理之前，本 实施例执行包括：对待拉伸区域进行低频区域提取处理，其中，低频区域为灰度值变化速率小于预设阈值的区域；根据调控拉伸参数，对待拉伸区域进行拉伸处理，包括：根据调控拉伸参数，对待拉伸区域的低频区域进行拉伸处理。提取待拉伸区域中的低频区域，并对低频区域进行拉伸处理，使得待拉伸区域中的高频区域或者噪声不被拉升，能够有效提高拉升后图像的信噪比，增加图像对比度，提高拉伸后图像的质量。对待拉伸区域进行低频区域提取处理的方法，可以是领域内常规的，例如可以利用小波变换的方法分理处待拉伸区域中的低频区域，此处不再赘述。
可选地，在根据调控拉伸参数，对待拉伸区域进行拉伸处理之前，本实施例还包括：对待拉伸区域进行细节提取处理，得到待拉伸区域的图像细节；在根据调控拉伸参数，对待拉伸区域进行拉伸处理之后，执行包括：将图像细节添加至拉伸处理后的图像中，得到处理后的图像。其中，图像细节包括待拉伸区域中占有少量像素的小目标或者目标表面的纹理信息，图像细节通常含有待拉伸区域的关键细节信息，但是图像细节的像素数较少，而且灰度值变化幅度较小，若强行对图像细节进行拉伸，可能导致图像细节失真，从而造成待拉伸区域的关键细节信息丢失，因此，本实施例中在对待拉伸区域进行拉伸处理前，先提取出待拉伸区域中的图像细节，并在对待拉伸区域进行拉伸处理后将提取出的图像细节重新添加至拉伸处理后的图像中。对待拉伸区域进行细节提取处理的方法可以是领域内常规的，例如，可以利用双边滤波器分离出待拉伸区域中的图像细节，在此不再赘述。
可选地，对待拉伸区域进行拉伸处理之后，本实施例还包括：对拉伸处理之后的图像进行平滑处理。其中，对拉伸处理后的图像进行平滑处理的方法可以是领域内常规的，例如，可以采用双边滤波器对局部拉伸后的图像进行边缘降噪的平滑处理，以降低局部拉伸后的图像的边缘光晕，达到保持边缘、降噪平滑的效果；或者，可以采用dither抖动算法对拉伸后的图像进行平滑处理，避免拉伸后的图像区域之间产生断阶。
可选地，本实施例的方法包括不同的拉伸模式，不同的拉伸模式对应于不同的拉伸调控特征。其中，拉伸模式包括高增益模式和低增益模式，或者低对比度模式和高对比度模式。
在一些可行的实施方式中，在对待拉伸区域进行灰度拉伸处理之后，将拉伸处理后的图像发送给显示设备，以供显示设备显示拉伸处理后的图像，显示设备可以包括设置在控制终端中的显示屏、显示器等。操作人员可以根据显示设备中显示的拉伸处理后的图像进行不同的应用，例如，根据显示设备中显示的拉伸处理后的图像进行目标追踪、红外测距、红外测温、远程遥控、实时监控等，示例性的，由设置在无人机中的红外相机拍摄基站的红外图像，由设置在无人机中的图像处理模块对红外图像进行灰度拉伸处理，并将拉伸处理后的图像发送给控制终端的显示设备，操作人员根据显示设备中显示的图像进行基站的温度判断，由此实现基站的红外远程测温。灰度拉伸处理后的图像对比度增强、图像细节明显、图像质量得以有效提升，根据拉伸处理后的图像能够实现更多种形式更高要求的红外图像应用。
本实施例中，通过确定待拉伸区域；确定待拉伸区域的灰度特征和拉伸调控特征；根据待拉伸区域的灰度特征，确定待拉伸区域的初始拉伸参数；根据拉伸调控特征，对初始拉伸参数进行调控，得到调控拉伸参数；根据调控拉伸参数，对待拉伸区域进行拉伸处理。在根据待拉伸区域的灰度特征，确定待拉伸区域的初始拉伸参数之前，对待拉伸区域中的图像坏点的灰度特征进行校正，有利于提高初始拉伸参数与待拉伸区域的匹配度；根据拉伸调控特征对初始拉伸参数进行多种形式的调控，得到的调控拉伸参数具有适用性强、灵活性高、匹配度完善的优点，待拉伸区域采用适配的调控拉伸参数进行灰度值拉伸处理，处理后的红外图像的图像质量高，图像细节改善明显，图像对比度增强效果显著。
图2a是本申请实施例提供的又一种图像处理方法，如图2a所示，包括：
S1、接收输入图像，检测输入图像的灰度值范围，并对输入图像中的图像坏点的灰度值进行校正；根据检测到的灰度值范围，确定第一初始拉伸参数；根据噪声强度对第一初始拉伸参数进行调控，得到第一调控拉伸参数；根据第一调控拉伸参数，对输入图像进行灰度拉伸处理，具体地，对输入图像进行线性拉伸处理。线性拉伸处理为以相同的拉伸强度和相同 的偏移幅度，对输入图像的所有像素点的灰度值进行拉伸处理，线性拉伸处理实现方式简单，拉伸效率高，有利于实现快速的图像灰度值优化处理。
S2、获取步骤S1处理后的输入图像的全局区域直方图，根据全局区域直方图确定全局区域的灰度值概率密度函数；根据全局区域的灰度值概率密度函数，确定全局区域的灰度值累积分布函数；根据全局区域的灰度值累积分布函数，确定第二初始拉伸参数；根据噪声强度对第二初始拉伸参数进行调控，得到第二调控拉伸参数；根据第二调控拉伸参数，对步骤S1处理后的输入图像进行全局区域的自适应拉伸处理。根据确定出的全局区域的灰度值累积分布函数，确定第二初始拉伸参数，第二初始拉伸参数与全局区域的像素点的灰度值具有适应性。在对第二初始拉伸参数进行调控后，对输入图像的全局区域进行灰度值拉伸处理，实现了对不同像素点的灰度值进行适应性的拉伸处理，灰度拉伸效果明显，有利于有效增加红外图像的对比度，和提高红外图像的图像质量。
S3、将步骤S2处理后的输入图像划分成多个部分区域，分别获取每个部分区域的部分区域直方图；分别根据每个部分区域直方图，确定每个部分区域的灰度值概率密度函数；根据每个部分区域的灰度值概率密度函数，确定每个部分区域的灰度值累积分布函数；根据每个部分区域的灰度值累积分布函数，确定每个部分区域的第三初始拉伸参数；根据每个部分区域的噪声强度，对每个部分区域的第三初始拉伸参数进行调控，得到每个部分区域的第三调控拉伸参数；根据每个部分区域的第三调控拉伸参数，分别对每个部分区域进行自适应拉伸处理。根据每个部分区域的灰度值累积分布函数，确定每个部分区域的第三初始拉伸参数，每个部分区域的第三初始拉伸参数与对应的部分区域的像素点灰度值具有适应性。在对每个第三初始拉伸参数进行调控后，根据每个部分区域的第三调控拉伸参数，分别对每个部分区域进行灰度拉伸处理，实现了对不同部分区域的不同像素点的灰度值进行适应性的拉伸处理，拉伸处理方式多样性丰富，灰度拉伸效果显著，红外图像的对比度得以有效提升，红外图像的图像质量进而也得以有效提高。
在对红外图像进行灰度拉伸处理时，可以同时采用S1、S2、S3三种灰度拉伸处理方法，也可以仅采用其中的一种或两种灰度拉伸处理方法， 并且不同灰度拉伸处理方法的使用次序可以任意调整。当同时采用不止一种方法进行灰度拉伸处理时，可以针对待拉伸区域的不同子区域采用不同的灰度拉伸处理方法，例如，可以根据不同的温度段将待拉伸区域划分为不同的子区域，对不同的子区域进行不同的灰度拉伸处理。对红外图像进行的灰度拉伸处理方式具有多样性和灵活性，在对红外图像进行增强处理时，可根据应用要求或用户需求进行拉伸处理方法的任意组合，有利于实现有效的红外图像对比度增强处理，有利于显著改善红外图像的图像质量。
图3为本申请实施例提供的一种图像处理装置的结构示意图，如图3所示，本申请实施例提供了一种图像处理装置，具体包括：处理器301和存储器302，其中，
存储器302，用于存储可被处理器执行的指令。
处理器301，用于执行存储器302中存储的指令，以实现图1-图2所示实施例中的图像处理方法。
具体的，处理器301执行指令以执行以下步骤：
确定待拉伸区域的灰度特征和拉伸调控特征，待拉伸区域对应输入图像的至少部分区域；
根据待拉伸区域的灰度特征，确定待拉伸区域的初始拉伸参数；
根据拉伸调控特征，对初始拉伸参数进行调控，得到调控拉伸参数；
根据调控拉伸参数，对待拉伸区域进行拉伸处理。
处理器301根据待拉伸区域的灰度特征，确定待拉伸区域的初始拉伸参数，包括以下至少一种：
根据待拉伸区域的灰度值范围，确定待拉伸区域的初始拉伸参数；
根据待拉伸区域的全局区域中灰度值与像素概率密度之间的映射关系，确定待拉伸区域的初始拉伸参数；
根据待拉伸区域的部分区域中灰度值与像素概率密度之间的映射关系，确定待拉伸区域的局部拉伸参数，局部拉伸参数构成初始拉伸参数，其中，局部拉伸参数与部分区域相对应。
处理器301在根据待拉伸区域的灰度值范围，确定待拉伸区域的初始拉伸参数之前，执行包括：
确定待拉伸区域中的图像坏点；
对图像坏点的灰度值进行校正处理。
具体的，拉伸调控特征包括以下至少一种：噪声、信噪比、或待拉伸区域的至少部分边缘。
当处理器301根据噪声对初始拉伸参数进行调控，得到调控拉伸参数时：噪声越大，调控拉伸参数对应的拉伸程度越小；
当处理器301根据信噪比对初始拉伸参数进行调控，得到调控拉伸参数时：信噪比越小，调控拉伸参数对应的拉伸程度越小；
当处理器301根据待拉伸区域的至少部分边缘对初始拉伸参数进行调控，得到调控拉伸参数时，得到的调控拉伸参数能够对待拉伸区域的至少部分边缘的拉伸程度进行抑制。
处理器301在根据待拉伸区域的部分区域中灰度值与像素概率密度之间的映射关系，确定待拉伸区域的局部拉伸参数之前，执行包括：确定部分区域，部分区域至少为两个，全部部分区域构成待拉伸区域；
处理器301根据待拉伸区域的部分区域中灰度值与像素概率密度之间的映射关系，确定待拉伸区域的局部拉伸参数，包括：分别根据待拉伸区域的各个部分区域中灰度值与像素概率密度之间的映射关系，确定各个部分区域所对应的局部拉伸参数。
处理器301根据拉伸调控特征，对初始拉伸参数进行调控，得到调控拉伸参数，包括：根据拉伸调控特征，分别对各个部分区域所对应的局部拉伸参数进行调控，得到各个部分区域所对应的调控拉伸参数；
处理器301根据调控拉伸参数，对待拉伸区域进行拉伸处理，包括：根据各个部分区域所对应的调控拉伸参数，对待拉伸区域的各个部分区域进行拉伸处理。
处理器301根据各个部分区域所对应的调控拉伸参数，对待拉伸区域的各个部分区域进行拉伸处理，还包括：
对于第一部分区域，根据与其相邻的部分区域所对应的调控拉伸参数对第一部分区域所对应的调控拉伸参数进行综合调整，其中，第一部分区域为部分区域中的任意一个；
根据综合调整后的各个部分区域所对应的调控拉伸参数，对待拉伸区 域的各个部分区域进行拉伸处理。
待拉伸区域至少为一个，处理器301在确定待拉伸区域的灰度特征之前，还执行包括：确定待拉伸区域，待拉伸区域可以根据输入的控制指令确定，或根据预先设置的程序确定。
输入的控制指令包括以下至少一种：
根据图像温度划分待拉伸区域；
根据图像中的物体划分待拉伸区域；
根据图像中的场景划分待拉伸区域。
处理器301在根据调控拉伸参数，对待拉伸区域进行拉伸处理之前，执行包括：
对待拉伸区域进行低频区域提取处理，其中，低频区域为灰度值变化速率小于预设阈值的区域；
处理器301根据调控拉伸参数，对待拉伸区域进行拉伸处理，包括：
根据调控拉伸参数，对待拉伸区域的低频区域进行拉伸处理。
在处理器301根据调控拉伸参数，对待拉伸区域进行拉伸处理之前，还执行包括：
对待拉伸区域进行细节提取处理，得到待拉伸区域的图像细节；
在处理器301根据调控拉伸参数，对待拉伸区域进行拉伸处理之后，执行包括：
将图像细节添加至拉伸处理后的图像中，得到处理后的图像。
处理器301在对待拉伸区域进行拉伸处理之后，还执行包括：对拉伸处理之后的图像进行平滑处理。
拉伸处理包括不同的拉伸模式，不同的拉伸模式对应于不同的拉伸调控特征。
处理器301在对待拉伸区域进行拉伸处理之后，执行包括：将拉伸处理后的图像发送给显示设备，显示设备用于显示拉伸处理后的图像。
可选地，图像处理装置还可以包括通信接口303和总线304，其中，处理器301、存储器302以及通信接口303可以通过总线304相互连接。通信接口303可以包括图像处理装置与红外相机，或与控制终端进行通信的接口。总线304可以是外设部件互连标准(Peripheral Component Interconnect， 简称PCI)总线或扩展工业标准结构(Extended Industry Standard Architecture，简称EISA)总线等。上述总线304可以分为地址总线、数据总线和控制总线等。为便于表示，图3中仅用一条粗线表示，但并不表示仅有一根总线或一种类型的总线。
在本申请实施例中，上述各实施例之间可以相互参考和借鉴，相同或相似的步骤以及名词均不再一一赘述。
或者，以上各个模块的部分或全部也可以通过集成电路的形式内嵌于该图像处理装置的某一个芯片上来实现。且它们可以单独实现，也可以集成在一起。即以上这些模块可以被配置成实施以上方法的一个或多个集成电路，例如：一个或多个特定集成电路(Application Specific Integrated Circuit，简称ASIC)，或，一个或多个微处理器(Digital Singnal Processor，简称DSP)，或，一个或者多个现场可编程门阵列(Field Programmable Gate Array，简称FPGA)。
本实施例中，通过确定待拉伸区域；确定待拉伸区域的灰度特征和拉伸调控特征；根据待拉伸区域的灰度特征，确定待拉伸区域的初始拉伸参数；根据拉伸调控特征，对初始拉伸参数进行调控，得到调控拉伸参数；根据调控拉伸参数，对待拉伸区域进行拉伸处理。在根据待拉伸区域的灰度特征，确定待拉伸区域的初始拉伸参数之前，对待拉伸区域中的图像坏点的灰度特征进行校正，有利于提高初始拉伸参数与待拉伸区域的匹配度；根据拉伸调控特征对初始拉伸参数进行多种形式的调控，得到的调控拉伸参数具有适用性强、灵活性高、匹配度完善的优点，待拉伸区域采用适配的调控拉伸参数进行灰度值拉伸处理，处理后的红外图像的图像质量高，图像细节改善明显，图像对比度增强效果显著。
图4为本申请实施例提供的一种图像处理系统的结构示意图，如图4所示，该图像处理系统包括红外相机和如图3所示的图像处理装置，其中，红外相机与图像处理装置对应连接，图像处理装置用于对红外相机采集的红外图像进行处理。
可选地，红外相机搭载于可移动平台上，可移动平台用于执行红外图像采集任务；红外相机用于在可移动平台执行红外图像采集任务时，采集 红外图像，并将红外图像发送给图像处理装置；图像处理装置用于接收红外相机拍摄的红外图像，并对红外图像进行处理。
可移动平台包括无人车、无人机、可移动机器人等。
可选地，图像处理系统还包括显示装置，显示装置用于显示图像处理装置处理后的图像。
可选地，图像处理还包括控制终端，控制终端可以向图像处理装置发送控制指令，控制指令用于确定待拉伸区域或确定图像拉伸模式。
可选地，显示装置设置在控制终端上。
一种计算机可读存储介质，其上存储有计算机程序，计算机程序被处理器执行以实现上述处理方法。
在上述实施例中，可以全部或部分地通过软件、硬件、固件或者其任意组合来实现。当使用软件实现时，可以全部或部分地以计算机程序产品的形式实现。计算机程序产品包括一个或多个计算机指令。在计算机上加载和执行计算机程序指令时，全部或部分地产生按照本申请实施例的流程或功能。计算机可以是通用计算机、专用计算机、计算机网络、或者其他可编程装置。计算机指令可以存储在计算机可读存储介质中，或者从一个计算机可读存储介质向另一个计算机可读存储介质传输，例如，计算机指令可以从一个网站站点、计算机、图像处理设备或数据中心通过有线(例如，同轴电缆、光纤、数字用户线(digital subscriber line，DSL))或无线(例如，红外、无线、微波等)方式向另一个网站站点、计算机、图像处理设备或数据中心进行传输。计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包含一个或多个可用介质集成的图像处理设备、数据中心等数据存储设备。可用介质可以是磁性介质，(例如，软盘、硬盘、磁带)、光介质(例如，DVD)、或者半导体介质(例如，固态硬盘(solid state disk，SSD))等。
本领域技术人员应该可以意识到，在上述一个或多个示例中，本申请实施例描述的功能可以用硬件、软件、固件或它们的任意组合来实现。当使用软件实现时，可以将这些功能存储在计算机可读介质中或者作为计算机可读介质上的一个或多个指令或代码进行传输。计算机可读介质包括计算机存储介质和通信介质，其中通信介质包括便于从一个地方向另一个地 方传送计算机程序的任何介质。存储介质可以是通用或专用计算机能够存取的任何可用介质。
本领域技术人员在考虑说明书及实践这里公开的发明后，将容易想到本公开的其它实施方案。本发明旨在涵盖本公开的任何变型、用途或者适应性变化，这些变型、用途或者适应性变化遵循本公开的一般性原理并包括本公开未公开的本技术领域中的公知常识或惯用技术手段。说明书和实施例仅被视为示例性的，本公开的真正范围和精神由下面的权利要求书指出。
应当理解的是，本公开并不局限于上面已经描述并在附图中示出的精确结构，并且可以在不脱离其范围进行各种修改和改变。本公开的范围仅由所附的权利要求书来限制。

Claims (36)

  1. 一种图像处理方法，其特征在于，所述方法用于红外图像的处理，所述方法包括：
    确定待拉伸区域的灰度特征和拉伸调控特征，所述待拉伸区域对应输入图像的至少部分区域；
    根据所述待拉伸区域的灰度特征，确定所述待拉伸区域的初始拉伸参数；
    根据所述拉伸调控特征，对所述初始拉伸参数进行调控，得到调控拉伸参数；
    根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理。
  2. 根据权利要求1所述的方法，其特征在于，所述根据所述待拉伸区域的灰度特征，确定所述待拉伸区域的初始拉伸参数，包括以下至少一种：
    根据所述待拉伸区域的灰度值范围，确定所述待拉伸区域的所述初始拉伸参数；
    根据所述待拉伸区域的全局区域中灰度值与像素概率密度之间的映射关系，确定所述待拉伸区域的所述初始拉伸参数；
    根据所述待拉伸区域的部分区域中灰度值与像素概率密度之间的映射关系，确定所述待拉伸区域的局部拉伸参数，所述局部拉伸参数构成所述初始拉伸参数，其中，所述局部拉伸参数与所述部分区域相对应。
  3. 根据权利要求2所述的方法，其特征在于，在根据所述待拉伸区域的灰度值范围，确定所述待拉伸区域的初始拉伸参数之前，执行包括：
    确定所述待拉伸区域中的图像坏点；
    对所述图像坏点的灰度值进行校正处理。
  4. 根据权利要求1-3任一项所述的方法，其特征在于，
    所述拉伸调控特征包括以下至少一种：噪声、信噪比、或所述待拉伸区域的至少部分边缘。
  5. 根据权利要求4所述的方法，其特征在于，
    当根据所述噪声对所述初始拉伸参数进行调控，得到所述调控拉伸参数时：所述噪声越大，所述调控拉伸参数对应的拉伸程度越小；
    当根据所述信噪比对所述初始拉伸参数进行调控，得到所述调控拉伸参数时：所述信噪比越小，所述调控拉伸参数对应的拉伸程度越小；
    当根据所述待拉伸区域的至少部分边缘对所述初始拉伸参数进行调控，得到所述调控拉伸参数时，得到的所述调控拉伸参数能够对所述待拉伸区域的至少部分边缘的拉伸程度进行抑制。
  6. 根据权利要求2所述的方法，其特征在于，所述根据所述待拉伸区域的部分区域中灰度值与像素概率密度之间的映射关系，确定所述待拉伸区域的局部拉伸参数之前，包括：确定所述部分区域，所述部分区域至少为两个，全部所述部分区域构成所述待拉伸区域；
    所述根据所述待拉伸区域的部分区域中灰度值与像素概率密度之间的映射关系，确定所述待拉伸区域的局部拉伸参数，包括：分别根据所述待拉伸区域的各个部分区域中灰度值与像素概率密度之间的映射关系，确定各个所述部分区域所对应的局部拉伸参数。
  7. 根据权利要求6所述的方法，其特征在于，所述根据所述拉伸调控特征，对所述初始拉伸参数进行调控，得到调控拉伸参数，包括：根据所述拉伸调控特征，分别对各个所述部分区域所对应的局部拉伸参数进行调控，得到各个所述部分区域所对应的调控拉伸参数；
    所述根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理，包括：根据各个所述部分区域所对应的调控拉伸参数，对所述待拉伸区域的各个所述部分区域进行拉伸处理。
  8. 根据权利要求6或7所述的方法，其特征在于，所述根据各个所述部分区域所对应的调控拉伸参数，对所述待拉伸区域的各个所述部分区域进行拉伸处理，还包括：
    对于第一部分区域，根据与其相邻的所述部分区域所对应的调控拉伸参数对所述第一部分区域所对应的调控拉伸参数进行综合调整，其中，所述第一部分区域为所述部分区域中的任意一个；
    根据所述综合调整后的各个所述部分区域所对应的调控拉伸参数，对所述待拉伸区域的各个所述部分区域进行拉伸处理。
  9. 根据权利要求1-8任意一项所述的方法，其特征在于，所述待拉伸区域至少为一个，所述确定待拉伸区域的灰度特征之前，还包括：确定 所述待拉伸区域，所述待拉伸区域可以根据输入的控制指令确定，或根据预先设置的程序确定。
  10. 根据权利要求9所述的方法，其特征在于，所述输入的控制指令包括以下至少一种：
    根据图像温度划分所述待拉伸区域；
    根据图像中的物体划分所述待拉伸区域；
    根据图像中的场景划分所述待拉伸区域。
  11. 根据权利要求1-10任一项所述的方法，其特征在于，在根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理之前，执行包括：
    对所述待拉伸区域进行低频区域提取处理，其中，所述低频区域为灰度值变化速率小于预设阈值的区域；
    所述根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理，包括：
    根据所述调控拉伸参数，对所述待拉伸区域的所述低频区域进行拉伸处理。
  12. 根据权利要求1-11任一项所述的方法，其特征在于，在根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理之前，还包括：
    对所述待拉伸区域进行细节提取处理，得到所述待拉伸区域的图像细节；
    在根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理之后，执行包括：
    将所述图像细节添加至拉伸处理后的图像中，得到所述处理后的图像。
  13. 根据权利要求1-12任一项所述的方法，其特征在于，所述对所述待拉伸区域进行拉伸处理之后，还包括：对所述拉伸处理之后的图像进行平滑处理。
  14. 根据权利要求1-13任一项所述的方法，其特征在于，所述方法包括不同的拉伸模式，所述不同的拉伸模式对应于不同的拉伸调控特征。
  15. 根据权利要求1-14任一项所述的方法，其特征在于，对所述待拉伸区域进行拉伸处理之后，还包括：将所述拉伸处理后的图像发送给显示设备，所述显示设备用于显示所述拉伸处理后的图像。
  16. 一种图像处理装置，其特征在于，所述装置用于红外图像的处理， 所述装置包括处理器和存储器，所述处理器与所述存储器通信连接，所述存储器存储有可被所述处理器执行的指令，所述处理器执行所述指令以执行以下步骤：
    确定待拉伸区域的灰度特征和拉伸调控特征，所述待拉伸区域对应输入图像的至少部分区域；
    根据所述待拉伸区域的灰度特征，确定所述待拉伸区域的初始拉伸参数；
    根据所述拉伸调控特征，对所述初始拉伸参数进行调控，得到调控拉伸参数；
    根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理。
  17. 根据权利要求16所述的装置，其特征在于，所述处理器根据所述待拉伸区域的灰度特征，确定所述待拉伸区域的初始拉伸参数，包括以下至少一种：
    根据所述待拉伸区域的灰度值范围，确定所述待拉伸区域的所述初始拉伸参数；
    根据所述待拉伸区域的全局区域中灰度值与像素概率密度之间的映射关系，确定所述待拉伸区域的所述初始拉伸参数；
    根据所述待拉伸区域的部分区域中灰度值与像素概率密度之间的映射关系，确定所述待拉伸区域的局部拉伸参数，所述局部拉伸参数构成所述初始拉伸参数，其中，所述局部拉伸参数与所述部分区域相对应。
  18. 根据权利要求17所述的装置，其特征在于，所述处理器在根据所述待拉伸区域的灰度值范围，确定所述待拉伸区域的初始拉伸参数之前，执行包括：
    确定所述待拉伸区域中的图像坏点；
    对所述图像坏点的灰度值进行校正处理。
  19. 根据权利要求16-18任一项所述的装置，其特征在于，
    所述拉伸调控特征包括以下至少一种：噪声、信噪比、或所述待拉伸区域的至少部分边缘。
  20. 根据权利要求19所述的装置，其特征在于，
    当所述处理器根据所述噪声对所述初始拉伸参数进行调控，得到所述 调控拉伸参数时：所述噪声越大，所述调控拉伸参数对应的拉伸程度越小；
    当所述处理器根据所述信噪比对所述初始拉伸参数进行调控，得到所述调控拉伸参数时：所述信噪比越小，所述调控拉伸参数对应的拉伸程度越小；
    当所述处理器根据所述待拉伸区域的至少部分边缘对所述初始拉伸参数进行调控，得到所述调控拉伸参数时，得到的所述调控拉伸参数能够对所述待拉伸区域的至少部分边缘的拉伸程度进行抑制。
  21. 根据权利要求17所述的装置，其特征在于，所述处理器在根据所述待拉伸区域的部分区域中灰度值与像素概率密度之间的映射关系，确定所述待拉伸区域的局部拉伸参数之前，执行包括：确定所述部分区域，所述部分区域至少为两个，全部所述部分区域构成所述待拉伸区域；
    所述处理器根据所述待拉伸区域的部分区域中灰度值与像素概率密度之间的映射关系，确定所述待拉伸区域的局部拉伸参数，包括：分别根据所述待拉伸区域的各个部分区域中灰度值与像素概率密度之间的映射关系，确定各个所述部分区域所对应的局部拉伸参数。
  22. 根据权利要求21所述的装置，其特征在于，所述处理器根据所述拉伸调控特征，对所述初始拉伸参数进行调控，得到调控拉伸参数，包括：根据所述拉伸调控特征，分别对各个所述部分区域所对应的局部拉伸参数进行调控，得到各个所述部分区域所对应的调控拉伸参数；
    所述处理器根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理，包括：根据各个所述部分区域所对应的调控拉伸参数，对所述待拉伸区域的各个所述部分区域进行拉伸处理。
  23. 根据权利要求21或22所述的装置，其特征在于，所述处理器根据各个所述部分区域所对应的调控拉伸参数，对所述待拉伸区域的各个所述部分区域进行拉伸处理，还包括：
    对于第一部分区域，根据与其相邻的所述部分区域所对应的调控拉伸参数对所述第一部分区域所对应的调控拉伸参数进行综合调整，其中，所述第一部分区域为所述部分区域中的任意一个；
    根据所述综合调整后的各个所述部分区域所对应的调控拉伸参数，对所述待拉伸区域的各个所述部分区域进行拉伸处理。
  24. 根据权利要求16-23任意一项所述的装置，其特征在于，所述待拉伸区域至少为一个，所述处理器在确定待拉伸区域的灰度特征之前，还执行包括：确定所述待拉伸区域，所述待拉伸区域可以根据输入的控制指令确定，或根据预先设置的程序确定。
  25. 根据权利要求24所述的装置，其特征在于，所述输入的控制指令包括以下至少一种：
    根据图像温度划分所述待拉伸区域；
    根据图像中的物体划分所述待拉伸区域；
    根据图像中的场景划分所述待拉伸区域。
  26. 根据权利要求16-25任一项所述的装置，其特征在于，所述处理器在根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理之前，执行包括：
    对所述待拉伸区域进行低频区域提取处理，其中，所述低频区域为灰度值变化速率小于预设阈值的区域；
    所述处理器根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理，包括：
    根据所述调控拉伸参数，对所述待拉伸区域的所述低频区域进行拉伸处理。
  27. 根据权利要求16-26任一项所述的装置，其特征在于，在所述处理器根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理之前，还执行包括：
    对所述待拉伸区域进行细节提取处理，得到所述待拉伸区域的图像细节；
    在所述处理器根据所述调控拉伸参数，对所述待拉伸区域进行拉伸处理之后，执行包括：
    将所述图像细节添加至拉伸处理后的图像中，得到所述处理后的图像。
  28. 根据权利要求16-27任一项所述的装置，其特征在于，所述处理器在对所述待拉伸区域进行拉伸处理之后，还执行包括：对所述拉伸处理之后的图像进行平滑处理。
  29. 根据权利要求16-28任一项所述的装置，其特征在于，所述拉伸 处理包括不同的拉伸模式，所述不同的拉伸模式对应于不同的拉伸调控特征。
  30. 根据权利要求16-19任一项所述的装置，其特征在于，所述处理器在对所述待拉伸区域进行拉伸处理之后，执行包括：将所述拉伸处理后的图像发送给所述显示设备，所述显示设备用于显示所述拉伸处理后的图像。
  31. 一种图像处理系统，其特征在于，包括红外相机和权利要求16所述的图像处理装置，其中，所述红外相机与所述图像处理装置连接，所述图像处理装置用于对所述红外相机采集的红外图像进行处理。
  32. 根据权利要求31所述的图像处理系统，其特征在于，所述红外相机搭载于可移动平台上，所述可移动平台用于执行红外图像采集任务；
    所述红外相机用于在所述可移动平台执行所述红外图像采集任务时，采集所述红外图像，并将所述红外图像发送给所述图像处理装置；
    所述图像处理装置用于接收所述红外相机拍摄的所述红外图像，并对所述红外图像进行处理。
  33. 根据权利要求31或32所述的图像处理系统，其特征在于，所述图像处理系统还包括显示装置，所述显示装置用于显示所述图像处理装置处理后的图像。
  34. 根据权利要求31-33任一项所述的图像处理系统，其特征在于，所述图像处理还包括控制终端，所述控制终端可以向所述图像处理装置发送控制指令，所述控制指令用于确定所述待拉伸区域或确定图像拉伸模式。
  35. 根据权利要求33或34所述的图像处理系统，其特征在于，所述显示装置设置在所述控制终端上。
  36. 一种储有计算机指令的非瞬时计算机可读存储介质，其特征在于，所述计算机指令用于使所述计算机执行权利要求1-15任一项所述的方法。
PCT/CN2020/087574 2020-04-28 2020-04-28 图像处理方法、装置、系统和存储介质 WO2021217445A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2020/087574 WO2021217445A1 (zh) 2020-04-28 2020-04-28 图像处理方法、装置、系统和存储介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2020/087574 WO2021217445A1 (zh) 2020-04-28 2020-04-28 图像处理方法、装置、系统和存储介质

Publications (1)

Publication Number Publication Date
WO2021217445A1 true WO2021217445A1 (zh) 2021-11-04

Family

ID=78331579

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/087574 WO2021217445A1 (zh) 2020-04-28 2020-04-28 图像处理方法、装置、系统和存储介质

Country Status (1)

Country Link
WO (1) WO2021217445A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116205910A (zh) * 2023-04-27 2023-06-02 四川省港奇电子有限公司 一种电源适配器用注塑温度自适应学习调控系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130028538A1 (en) * 2011-07-29 2013-01-31 Simske Steven J Method and system for image upscaling
CN105427255A (zh) * 2015-11-16 2016-03-23 中国航天时代电子公司 一种基于grhp的无人机红外图像细节增强方法
CN110800309A (zh) * 2018-08-23 2020-02-14 深圳市大疆创新科技有限公司 图像处理方法、图像处理装置、电子装置和存储介质
CN110869976A (zh) * 2018-12-04 2020-03-06 深圳市大疆创新科技有限公司 一种图像处理方法、设备、无人机、系统及存储介质
CN111047624A (zh) * 2019-12-27 2020-04-21 成都英飞睿技术有限公司 图像弱小目标检测方法、装置、设备及存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130028538A1 (en) * 2011-07-29 2013-01-31 Simske Steven J Method and system for image upscaling
CN105427255A (zh) * 2015-11-16 2016-03-23 中国航天时代电子公司 一种基于grhp的无人机红外图像细节增强方法
CN110800309A (zh) * 2018-08-23 2020-02-14 深圳市大疆创新科技有限公司 图像处理方法、图像处理装置、电子装置和存储介质
CN110869976A (zh) * 2018-12-04 2020-03-06 深圳市大疆创新科技有限公司 一种图像处理方法、设备、无人机、系统及存储介质
CN111047624A (zh) * 2019-12-27 2020-04-21 成都英飞睿技术有限公司 图像弱小目标检测方法、装置、设备及存储介质

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116205910A (zh) * 2023-04-27 2023-06-02 四川省港奇电子有限公司 一种电源适配器用注塑温度自适应学习调控系统

Similar Documents

Publication Publication Date Title
US9842382B2 (en) Method and device for removing haze in single image
CN104978710B (zh) 一种基于拍照识别调节人脸亮度的方法和装置
US9251573B2 (en) Device, method, and storage medium for high dynamic range imaging of a combined image having a moving object
WO2017076040A1 (zh) 应用于连拍过程中的图像处理方法和装置
US10038862B2 (en) Methods and apparatus for automated noise and texture optimization of digital image sensors
US8917938B2 (en) Method for estimating blur degree of image and method for evaluating image quality
CN108965839B (zh) 一种自动调整投影画面的方法及装置
CN105096267B (zh) 一种基于拍照识别调节眼部亮度的方法和装置
US9478016B2 (en) Image restoration method and image processing apparatus using the same
US9396526B2 (en) Method for improving image quality
WO2021218603A1 (zh) 图像处理方法及投影系统
CN112116542B (zh) 图像对比度增强方法、装置、电子设备和存储介质
CN110736610A (zh) 测量光学中心偏移的方法、装置、存储介质及深度相机
JP2015154102A (ja) 画像処理装置及び方法、画像処理プログラム、撮像装置
US8942477B2 (en) Image processing apparatus, image processing method, and program
US20190273857A1 (en) Image pickup apparatus, image pickup method, and recording medium
CN113449730A (zh) 图像处理方法、系统、自动行走设备及可读存储介质
WO2021128498A1 (zh) 图像自适应降噪方法及装置
CN113379609B (zh) 一种图像处理方法、存储介质及终端设备
US9571744B2 (en) Video processing method and apparatus
WO2021217445A1 (zh) 图像处理方法、装置、系统和存储介质
CN114998122A (zh) 一种低照度图像增强方法
CN114125280B (zh) 相机曝光控制方法、装置、设备、存储介质
CN106847150B (zh) 调节显示屏亮度的装置及方法
CN110136085B (zh) 一种图像的降噪方法及装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20933667

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20933667

Country of ref document: EP

Kind code of ref document: A1