Infrared image enhancement method based on adaptive histogram equalization and improved Laplacian sharpening

Abstract

The invention provides an infrared image enhancement method based on adaptive histogram equalization and improved Laplacian sharpening. The infrared image enhancement method comprises the following steps: step 1, acquiring original data of an infrared image; step 2, carrying out bilateral filtering processing; step 3, histogram equalization processing is carried out, and adaptive histogram equalization infrared image information is obtained; step 4, Gaussian filtering processing is carried out; step 5, carrying out weight distribution on the Laplacian detail image to obtain a weight infrared image; step 6, summing the adaptive histogram equalization infrared image information and the weight infrared image to obtain an infrared enhanced image; according to the infrared image enhancement method based on adaptive histogram equalization and improved Laplacian sharpening, the contrast and sharpness of low-frequency information of the infrared image can be enhanced; therefore, the defects of low signal-to-noise ratio, high spatial correlation, low contrast ratio, fuzzy visual effect and the like of the infrared image are overcome, and the effects of tasks such as target detection tracking and the like are improved.

Description

Technical Field

[0001] The invention belongs to the technical field of infrared image processing, and in particular relates to an infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening. Background Art

[0002] With the rapid development of infrared imaging technology, it is now widely used in various fields. Infrared imaging technology has the advantages of real-time observation, but because infrared imaging is affected by external environmental interference, transmission distance and small radiation dynamic range, compared with visible light images, infrared images have low contrast, blurred image details and excessive noise.

[0003] Infrared image enhancement is an important step in the visualization process. However, traditional image enhancement algorithms will cause excessive enhancement of the background area of ​​the image when processing infrared images, ignore the enhancement of the detail area, and fail to obtain the ideal infrared image enhancement effect.

[0004] Existing infrared image enhancement algorithms include infrared image enhancement methods using histogram equalization. These infrared image enhancement methods have certain infrared image enhancement effects and improve image contrast. However, in pedestrian environments with brighter backgrounds, the enhancement effect is not soft and is rather harsh, making it impossible to distinguish pedestrians from the background.

[0005] The main reasons are the uninterrupted heat transfer and thermal radiation between the target and the background, the relatively long wavelength of infrared light and other inherent characteristics, as well as the long transmission distance of infrared radiation and the influence of atmospheric attenuation. These factors result in low signal-to-noise ratio, strong spatial correlation, low contrast, and blurred visual effects in infrared images, which have a great impact on tasks such as target detection and tracking. Summary of the invention

[0006] In order to solve the inherent characteristics of uninterrupted heat transfer and heat radiation between the target and the background, the relatively long wavelength of infrared light, and the influence of factors such as the long transmission distance of infrared radiation and atmospheric attenuation, the infrared image has low signal-to-noise ratio, strong spatial correlation, low contrast, and blurred visual effects, which has a great impact on tasks such as target detection and tracking. Therefore, the adaptive histogram equalization module designed in the present invention enhances the contrast of low-frequency information of infrared images, and designs an improved Laplace sharpening algorithm to enhance the sharpness of high-frequency information of infrared images.

[0007] To this end, the present invention provides an infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening, comprising the following steps:

[0008] Step 1: Get the original infrared image data org ;

[0009] Step 2: Raw infrared image data org , perform bilateral filtering to obtain the infrared image low-frequency information data base ;

[0010] Step 3: Use the adaptive histogram equalization module to equalize the infrared image low-frequency information data base Perform histogram equalization to obtain adaptive histogram equalization infrared image information data lp ;

[0011] Step 4: Red infrared image raw data org , perform Gaussian filtering to obtain Laplace detail infrared image data detail_lap ;

[0012] Step 5: Laplace detail image data detail_lap , perform weight assignment and obtain weighted infrared image data detail_Q ;

[0013] Step 6: Adaptively equalize the infrared image information data using histogram lp and weighted infrared image data detail_Q Perform weighted summation to obtain infrared enhanced image data detail_result .

[0014] Furthermore, in step 2, the infrared image data org , perform bilateral filtering to obtain infrared low-frequency image data base The calculation formula is:

[0015] data base =Bilt_gray(data org ) (1)

[0016] In the above formula, data org represents the received infrared image raw data, Bilt_gray(·) represents the bilateral filtering operation, data base It represents the low-frequency information of the infrared image data obtained after the original infrared image data is subjected to bilateral filtering.

[0017] Furthermore, in step 3, an adaptive histogram equalization module is used to adjust the infrared image low-frequency information data base Perform histogram equalization to obtain adaptive histogram equalization infrared image information data lp The specific process is:

[0018] First, the data base Perform truncation operation on data baseSort from small to large, sum represents data base The number of elements, denoted by clip low and clip high They are data base The sorted 3277th data and sum-3277 data;

[0019]

[0020] Where m and n represent the image length and image width respectively, (i, j) represents the image element index is the i-th row and j-th column, and formula (2) represents the low-frequency information data base Perform truncation operation and perform histogram equalization processing on the truncated data with adaptive mapping range calculation; use clip low and clip high Find the mapping range scale scaling factor θ;

[0021]

[0022] Contrast=θ·ρ (4)

[0023] Among them, θ is the scaling factor of the mapping range, Contrast is the obtained mapping range, ε and ρ are the default parameters, ε = 1700, ρ = 170, and after Contrast is obtained according to the original data, it is necessary to perform a protection operation of 70≤Contrast≤220 on it;

[0024] The parameters obtained in the above formula are used to linearly map the original data and then perform histogram equalization on it. The formula is as follows:

[0025]

[0026] Among them, HIST(·) represents the histogram equalization operation, data lp This is the output of the adaptive histogram equalization module we obtained.

[0027] Furthermore, in step 4, the original data of the red infrared image data org , perform Gaussian filtering to obtain Laplace detail infrared image data detail_lap The calculation process is:

[0028] data gauss =Gaussfilter(data org ) (6)

[0029] data detail_lap = Laplace(datagauss ) (7)

[0030] Among them, data gauss represents the result of Gaussian filtering of the original data. Gaussfilter(·) represents the Gaussian filtering operation. data detail_lap Represents the obtained Laplace detail infrared image.

[0031] Furthermore, in step 5, the Laplace detail image data detail_lap , perform weight assignment and obtain weighted infrared image data detail_Q The calculation formula is:

[0032]

[0033] Among them, data detail_lap (i, j) represents the element value of the i-th row and j-th column of the Laplace infrared detail image, and the value of φ is set to 200; in the above formula Represents element-by-element multiplication, for the detail layer data obtained in the above formula detail_Q The value of is truncated to the range of -255 to 255, and the truncated detail layer is linearly mapped in the range of -40 to 40.

[0034] Furthermore, in step 6, the infrared image information data is adaptively histogram equalized lp and weighted infrared image data detail_Q Perform weighted summation, α is the detail enhancement factor (0.2<α<2), and obtain the infrared enhanced image data detail_result The calculation formula is:

[0035] data detail_result =α*data detail_Q +data lp (10).

[0036] The advantages of the present invention are as follows: the present invention provides an infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening, the designed adaptive histogram equalization module can enhance the contrast of low-frequency information of infrared images, and the designed improved Laplace sharpening algorithm can enhance the sharpness of high-frequency information of infrared images; thereby improving the shortcomings of infrared images such as low signal-to-noise ratio, strong spatial correlation, low contrast, and blurred visual effects, and improving the effects of tasks such as target detection and tracking.

[0037] The present invention is described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1The figure is a flow chart of the infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening.

[0039] Figure 2 Schematic diagram of the dimming layer image output by the infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening.

[0040] Figure 3 Schematic diagram of the final image output by the infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening. DETAILED DESCRIPTION

[0041] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined purpose, the specific implementation methods, structural features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and embodiments.

[0042] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0043] All features disclosed in this specification, or steps in all methods or processes disclosed, except mutually exclusive features and / or steps, can be combined in any manner.

[0044] Any feature disclosed in this specification (including any additional claims, abstract and drawings), unless otherwise stated, may be replaced by other equivalent or alternative features with similar purposes. That is, unless otherwise stated, each feature is only an example of a series of equivalent or similar features.

[0045] Example 1

[0046] This embodiment proposes a Figure 1 The infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening shown in the figure comprises the following steps:

[0047] Step 1: Get the original infrared image data org ;

[0048] Step 2: Raw infrared image data org , perform bilateral filtering to obtain the infrared image low-frequency information data base ;

[0049] Step 3: Use the adaptive histogram equalization module to equalize the infrared image low-frequency information database Perform histogram equalization to obtain adaptive histogram equalization infrared image information data lp ;

[0050] Step 4: Red infrared image raw data org , perform Gaussian filtering to obtain Laplace detail infrared image data detail_lap ;

[0051] Step 5: Laplace detail image data detail_lap , perform weight assignment and obtain weighted infrared image data detail_Q ;

[0052] Step 6: Adaptively equalize the infrared image information data using histogram lp and weighted infrared image data detail_Q Perform weighted summation to obtain infrared enhanced image data detail_result .

[0053] Furthermore, in step 2, the infrared image data org , perform bilateral filtering to obtain infrared low-frequency image data base The calculation formula is:

[0054] data base =Bilt_gray(data org ) (1)

[0055] In the above formula, data org represents the received infrared image raw data, Bilt_gray(·) represents the bilateral filtering operation, data base It represents the low-frequency information of the infrared image data obtained after the original infrared image data is subjected to bilateral filtering.

[0056] Furthermore, in step 3, an adaptive histogram equalization module is used to adjust the infrared image low-frequency information data base Perform histogram equalization to obtain adaptive histogram equalization infrared image information data lp The specific process is:

[0057] First, the data base Perform truncation operation on data base Sort from small to large, sum represents data base The number of elements, denoted by clip low and clip high They are data base The sorted 3277th data and sum-3277 data;

[0058]

[0059] Where m and n represent the image length and image width respectively, (i, j) represents the image element index is the i-th row and j-th column, and formula (2) represents the low-frequency information data base Perform truncation operation and perform histogram equalization processing on the truncated data with adaptive mapping range calculation; use clip low and clip high Find the mapping range scale scaling factor θ;

[0060]

[0061] Contrast=θ·ρ (4)

[0062] Among them, θ is the scaling factor of the mapping range, Contrast is the obtained mapping range, ε and ρ are the default parameters, ε = 1700, ρ = 170, and after Contrast is obtained according to the original data, it is necessary to perform a protection operation of 70≤Contrast≤220 on it;

[0063] The parameters obtained in the above formula are used to linearly map the original data and then perform histogram equalization on it. The formula is as follows:

[0064]

[0065] Among them, HIST(·) represents the histogram equalization operation, data lp This is the output of the adaptive histogram equalization module we obtained.

[0066] Furthermore, in step 4, the original data of the red infrared image data org , perform Gaussian filtering to obtain Laplace detail infrared image data detail_lap The calculation process is:

[0067] data gauss =Gaussfilter(data org ) (6)

[0068] data detail_lap = Laplace(data gauss ) (7)

[0069] Among them, data gauss represents the result of Gaussian filtering of the original data. Gaussfilter(·) represents the Gaussian filtering operation. data detail_lap Represents the obtained Laplace detail infrared image.

[0070] Furthermore, in step 5, the Laplace detail image data detail_lap , perform weight assignment and obtain weighted infrared image data detail_Q The calculation formula is:

[0071]

[0072] Among them, data detail_lap (i, j) represents the element value of the i-th row and j-th column of the Laplace infrared detail image, and the value of φ is set to 200; in the above formula Represents element-by-element multiplication, for the detail layer data obtained in the above formula detail_Q The value of is truncated to the range of -255 to 255, and the truncated detail layer is linearly mapped in the range of -40 to 40.

[0073] Furthermore, in step 6, the infrared image information data is adaptively histogram equalized lp and weighted infrared image data detail_Q Perform weighted summation, α is the detail enhancement factor (0.2<α<2), and obtain the infrared enhanced image data detail_result The calculation formula is:

[0074] data detail_result =α*data detail_Q +data lp (10).

[0075] In summary, this embodiment provides an infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening. The designed adaptive histogram equalization module can enhance the contrast of low-frequency information of infrared images, and the designed improved Laplace sharpening algorithm can enhance the sharpness of high-frequency information of infrared images; thereby improving the shortcomings of infrared images such as low signal-to-noise ratio, strong spatial correlation, low contrast, and blurred visual effects, and improving the effects of tasks such as target detection and tracking.

[0076] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0077] Example 2

[0078] The infrared image is processed using the infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening shown in Example 1; Figure 2 The figure shows the dimming layer image effect obtained by the infrared image through steps 1 to 3 of this method. The low-frequency dimming layer image obtained by bilateral filtering and adaptive histogram equalization presents the advantages of high overall contrast, strong transparency and low noise level. Figure 3The infrared enhanced image obtained through steps 4 to 6 of the method shown in the figure further improves the sharpness and clarity of the high-frequency information of the infrared image by using the improved Laplace sharpening enhancement algorithm. In this image, the edges of the building and the windows on the surface are clearer without significantly increasing the overall noise of the image. This method improves the shortcomings of infrared images such as low signal-to-noise ratio, strong spatial correlation, low contrast, and blurred visual effects.

[0079] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.

Claims

1. An infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening, characterized in that: The steps include: Step 1: Get the original infrared image data org ; Step 2: Raw infrared image data org , perform bilateral filtering to obtain the infrared image low-frequency information data base ; Step 3: Use the adaptive histogram equalization module to equalize the infrared image low-frequency information data base Perform histogram equalization to obtain adaptive histogram equalization infrared image information data lp ; Step 4: Red infrared image raw data org , perform Gaussian filtering to obtain Laplace detail infrared image data detail_lap ; Step 5: Laplace detail image data detail_lap , perform weight assignment and obtain weighted infrared image data detail_Q ; Step 6: Adaptively equalize the infrared image information data using histogram lp and weighted infrared image data detail_Q Perform weighted summation to obtain infrared enhanced image data detail_result .

2. The infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening as claimed in claim 1, characterized in that: Step 2: org , perform bilateral filtering to obtain infrared low-frequency image data base The calculation formula is: data base =Bilt_gray(data org ) (1) In the above formula, data org represents the received infrared image raw data, Bilt_gray(·) represents the bilateral filtering operation, data base It represents the low-frequency information of the infrared image data obtained after the original infrared image data is subjected to bilateral filtering.

3. The infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening as claimed in claim 1, characterized in that: Step 3: Use the adaptive histogram equalization module to adjust the infrared image low-frequency information data base Perform histogram equalization to obtain adaptive histogram equalization infrared image information data lp The specific process is: First, the data base Perform truncation operation on data base Sort from small to large, sum represents data base The number of elements, denoted by clip low and clip high They are data base The sorted 3277th data and sum-3277 data; Where m and n represent the image length and image width respectively, (i, j) represents the image element index is the i-th row and j-th column, and formula (2) represents the low-frequency information data base Perform truncation operation and perform histogram equalization processing on the truncated data with adaptive mapping range calculation; use clip low and clip high Find the mapping range scale scaling factor θ; Contrast=θ·ρ (4) Among them, θ is the scaling factor of the mapping range, Contrast is the obtained mapping range, ε and ρ are the default parameters, ε = 1700, ρ = 170, and after Contrast is obtained according to the original data, it is necessary to perform a protection operation of 70≤Contrast≤220 on it; The parameters obtained in the above formula are used to linearly map the original data and then perform histogram equalization on it. The formula is as follows: Among them, HIST(·) represents the histogram equalization operation, data lp This is the output of the adaptive histogram equalization module we obtained.

4. The infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening as claimed in claim 1, characterized in that: In step 4, the original data of the red infrared image is org , perform Gaussian filtering to obtain Laplace detail infrared image data detail_lap The calculation process is: data gauss =Gaussfilter(data org ) (6) data detail_lap =Laplace(data gauss ) (7) Among them, data gauss represents the result of Gaussian filtering of the original data. Gaussfilter(·) represents the Gaussian filtering operation. data detail_lap Represents the obtained Laplace detail infrared image.

5. The infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening as claimed in claim 1, characterized in that: Step 5: Laplace detail image data detail_lap , perform weight assignment and obtain weighted infrared image data detail_Q The calculation formula is: Among them, data detail_lap (i, j) represents the element value of the i-th row and j-th column of the Laplace infrared detail image, and the value of φ is set to 200; in the above formula Represents element-by-element multiplication, for the detail layer data obtained in the above formula detail_Q The value of is truncated to the range of -255 to 255, and the truncated detail layer is linearly mapped in the range of -40 to 40.

6. The infrared image enhancement method based on adaptive histogram equalization and improved Laplace sharpening as claimed in claim 1, characterized in that: Step 6: Adaptively histogram equalize the infrared image information data lp and weighted infrared image data detail_Q Perform weighted summation, α is the detail enhancement factor (0.2<α<2), and obtain the infrared enhanced image data detail_result The calculation formula is: data detail_result =α*data detail_Q +data lp (10)。

Images