CN110717862B - Contrast enhancement method based on dynamic range compression and electronic device thereof - Google Patents

Abstract

The invention provides a contrast enhancement method based on dynamic range compression and an electronic device thereof, which determine a proper effective dynamic range through the occurrence times of an input brightness value of each pixel position in an input image, estimate a global mapping curve according to the effective dynamic range, and then adjust an output brightness value of an input brightness value mapping according to the regional characteristics of each pixel position in the input image so as to adaptively improve the contrast of the input image. Accordingly, the contrast enhancement method and the electronic device thereof can reduce the calculation complexity and generate better image contrast.

Description

Contrast enhancement method based on dynamic range compression and electronic device thereof

technical field

The present invention provides a contrast enhancement method and an electronic device thereof, in particular to a contrast enhancement method based on dynamic range compression and an electronic device thereof.

Background technique

High dynamic range (eg 32-bit) images can capture images that look like real scenes, because they retain a lot of the brightness, contrast, image details, and so on of the real scene. The dynamic range that a general display can display is, for example, 0 to 255 (ie, 8 bits). In order to allow general monitors to present high dynamic range images, high dynamic range images are enhanced by a compression method and appropriate image contrast to be close to the perception of human vision.

Image contrast enhancement can be divided into global (global) contrast enhancement and regional (local) contrast enhancement. Global contrast enhancement (such as gamma correction or histogram equalization) is to estimate a global curve with dynamic range compression and map the image. Although global contrast enhancement can process images quickly, such methods can over-compress too bright pixels or over-amplify too dark pixels, which often loses the contrast of the resulting image. Regional contrast enhancement (such as adaptive histogram equalization, exposure and shading techniques (Dodging-and-Burning)) is based on the relationship between each pixel and its neighbors to generate a nonlinear curve to adjust each pixel. Although regional contrast enhancement can produce better image contrast, the computational complexity of such methods is very high.

Therefore, if the advantages of the above-mentioned global contrast enhancement and regional contrast enhancement can be combined, the computational complexity can be reduced and better image contrast can be generated.

SUMMARY OF THE INVENTION

The present invention provides a contrast enhancement method based on dynamic range compression and an electronic device thereof, which determine an appropriate effective dynamic range by the number of occurrences of the input luminance value of each pixel position in an input image, and estimate the global map accordingly. Curve (Global Mapping Curve), and then adjust the output brightness value of the input brightness value mapping according to the regional characteristics of each pixel position in the input image, so as to adaptively improve the contrast of the input image. Accordingly, the contrast enhancement method and the electronic device thereof of the present invention can reduce computational complexity and generate better image contrast.

Embodiments of the present invention provide a contrast enhancement method based on dynamic range compression, which is suitable for an electronic device and used to adjust an input brightness of each pixel position in an input image to enhance the contrast of the input image. The contrast enhancement method includes the following steps: (A) receiving each input brightness in the input image; (B) corresponding an occurrence number of each input brightness value to a plurality of brightness values on the histogram, and performing these occurrence times to Smooth filtering, and determine an effective dynamic range according to the occurrences of smoothing; (C) sequentially accumulating each occurrence of smoothing in the effective dynamic range to generate an accumulation curve, wherein the accumulation curve represents the brightness values and the accumulation (D) normalize the accumulated occurrence times to the effective dynamic range to generate an output luminance value, and these luminance values and the output luminance value corresponding to each luminance value form a global mapping curve (E) in the global mapping curve, obtain the corresponding output luminance value according to the input luminance value of each pixel position in sequence; and (F) in each pixel position, according to the corresponding output luminance value and adjacent these outputs A brightness relationship between the brightness values adjusts the corresponding output brightness value to generate a final brightness value.

Embodiments of the present invention provide an electronic device based on dynamic range compression for adjusting an input brightness of each pixel position in an input image to enhance the contrast of the input image. The electronic device includes an image acquisition device and an image processor. The image acquisition device receives the input image, and sequentially acquires each input brightness in the input image. The image processor is electrically connected to the image acquisition device, and is used for performing the following steps: (A) receiving each input luminance value in the input image; (B) corresponding to a number of occurrences of each input luminance to multiple values on the histogram A luminance value is obtained, these occurrences are smoothed and filtered, and an effective dynamic range is determined according to the smoothed occurrences; (C) each smoothed occurrence in the effective dynamic range is sequentially accumulated to generate an accumulation curve, The accumulation curve represents the relationship between these brightness values and the accumulated occurrences; (D) normalize the accumulated occurrences to an effective dynamic range to generate an output brightness value, and these brightness values correspond to each brightness value (E) in the global mapping curve, obtain the corresponding output luminance value according to the input luminance value of each pixel position in sequence; and (F) in each pixel position, according to the corresponding A brightness relationship between the output brightness value of and the adjacent output brightness values adjusts the corresponding output brightness value to generate a final brightness value.

In order to further understand the features and technical content of the present invention, please refer to the following detailed descriptions and drawings related to the present invention, but these descriptions and drawings are only used to illustrate the present invention, rather than make any claims of the present invention. limits.

Description of drawings

FIG. 1 is a schematic diagram of an electronic device based on dynamic range compression according to an embodiment of the present invention.

FIG. 2 is a flowchart of a contrast enhancement method based on dynamic range compression according to an embodiment of the present invention.

FIG. 2A is a detailed flowchart of step S230 of FIG. 2 .

FIG. 2B is a detailed flowchart of step S270 of FIG. 2 .

FIG. 3 is a histogram of an input image according to an embodiment of the present invention.

FIG. 4 is a smoothed histogram of FIG. 3 .

FIG. 5 is a schematic diagram of an accumulation curve according to an embodiment of the present invention.

FIG. 6 is a schematic diagram of a global mapping curve according to an embodiment of the present invention.

FIG. 7 is a schematic diagram of an output luminance value of a current pixel position and an adjacent output luminance value according to an embodiment of the present invention.

FIG. 8 is a schematic diagram of adjusting the output luminance value of the current pixel position according to an embodiment of the present invention.

FIG. 9 is a schematic diagram of adjusting the output luminance value of the current pixel position according to another embodiment of the present invention.

FIG. 10 is a schematic diagram of adjusting the output luminance value of the current pixel position according to another embodiment of the present invention.

Symbol Description

100: Electronics

110: Image acquisition device

120: Image processor

Fr, Fr1: Input image

P0-Pn: Input brightness value

P0'-Pn': last brightness value

S210, S220, S230, S240, S250, S260, S270: Steps

S231, S233, S235, S237, S238, S239: Steps

S271, S273: Steps

L0: The first luminance value

L1: The first effective luminance value

L2: Second effective luminance value

DR1: first range

DR2: Second range

Deff: effective dynamic range

Cuv: cumulative curve

F1, F2, F3: Pixel groups

P22, P47, P102: Pixel position

Detailed ways

Hereinafter, the present invention will be described in detail by illustrating various exemplary embodiments of the invention by means of the accompanying drawings. However, the inventive concepts may be embodied in many different forms and should not be construed as limited to the exemplary embodiments set forth herein. Furthermore, the same reference numbers may be used in the figures to refer to similar elements.

First, please refer to FIG. 1 , which shows a schematic diagram of an electronic device based on dynamic range compression according to an embodiment of the present invention. As shown in FIG. 1 , the electronic device 100 is used to adjust the input luminance values P0-Pn of each pixel position in an input image Fr, so as to enhance the contrast of the input image Fr, and output the adjusted final luminance values P0'-Pn. Pn'. In this embodiment, the electronic device 100 may be a smart phone, a video recorder, a tablet computer, a notebook computer, or other devices that need to perform image contrast enhancement, which is not limited in the present invention.

The electronic device 100 includes an image acquisition device 110 and an image processor 120 . As shown in FIG. 1 , the image acquisition device 110 receives the input image Fr, and sequentially acquires each input luminance value P0-Pn in the input image Fr. More specifically, the image acquisition device 110 acquires continuous images, and the input image Fr is one of the continuous images. And each pixel position in the input image Fr has input luminance values P0-Pn respectively.

The image processor 120 is electrically connected to the image acquisition device 110, and is used for executing the following steps to adjust the input luminance values P0-Pn of each pixel position in the input image Fr, thereby enhancing the contrast of the input image Fr.

Please also refer to Figure 1 to Figure 2. FIG. 2 shows a flowchart of a contrast enhancement method based on dynamic range compression according to an embodiment of the present invention. First, the image processor 120 receives the input luminance values P0-Pn of each pixel position in the input image Fr to further analyze the characteristics of each input luminance value P0-Pn in the input image Fr (step S210).

Next, the image processor 120 corresponds an occurrence of each input luminance value P0-Pn to a plurality of luminance values on the histogram (step S220). As shown in FIG. 3 , the dynamic range of the luminance value of the histogram is 9 bits (bit, bit), that is, the luminance value is 0-511. Therefore, the image processor 120 counts the number of occurrences H(n) of each input luminance value P0-Pn in the input image Fr to the luminance value of the histogram. In this embodiment, the number of occurrences of the luminance value 0 is 10 times (represented by H(0)=10). In other luminance values 1-511, H(1)=15; H(2)=12; H(3)=8; H(4)=15; H(5)-H(10)=10; H(11)-H(248)=0; H(249)=10; H(250)=10; H(251)=5; H(252)=0; H(253)=1; H(254) )=0; H(255)=1; H(256)=1; H(257)=0; H(258)=1; H(259)=1; and H(260)-H(511)= 0. The dynamic range of the luminance value of the histogram can also be designed according to the actual situation, which is not limited in the present invention.

After obtaining the histogram, the image processor 120 performs smooth filtering on the occurrences, and determines an effective dynamic range according to the smoothed occurrences (step S230 ). It is worth noting that if the effective dynamic range is too small, the brighter parts of the input image Fr will be overexposed on the output image; on the contrary, if the effective dynamic range is too large, the darker parts of the input image Fr will appear. Too dark on the output graph. Therefore, an appropriate effective dynamic range will result in a better output result map.

Further, referring to FIG. 2A at the same time, the image processor 120 searches the histogram from the last luminance value forward to the first luminance value with the number of occurrences as a first effective luminance value, and uses the first luminance value as a first effective luminance value. One luminance value to the first effective luminance value is used as a first range (step S231). Taking FIG. 3 as an example, the image processor 120 searches forward from the last luminance value 511 for the first luminance value 259 that has the number of occurrences, and takes the luminance value 259 as the first effective luminance value L1 . The image processor 120 then uses the first luminance value L0 to the first effective luminance value L1 as a first range DR1.

Next, in the first range, the image processor 120 performs smooth filtering on the corresponding occurrence times to generate a smoothed histogram (step S233 ). In this embodiment, as shown in FIG. 3 , the image processor 120 uses a linear filter to perform smooth filtering on the occurrence times H(0)-H(259) in the first range DR1 to generate smoothing The smoothed occurrence times H'(n), and other methods can also be used to filter these occurrence times H(0)-H(259) to generate the smoothed histogram Histogram1 shown in FIG. No restrictions apply.

Please also refer to FIG. 4 , following the above example, the image processor 120 averages the number of adjacent occurrences according to the order of the luminance values. Therefore, the number of occurrences of smoothing is H'(0)=(10+15)/2=13 and H'(1)=(10+15+12)/3=12. The calculation method of other smoothed occurrences H'(2)-H'(510) is roughly the same as that of H'(1), and the calculation method of H'(511) is roughly the same as that of H'(0) The calculation method is the same, and the calculation result is shown in Figure 4, so it is not repeated here. Of course, N (N is a positive integer) luminance values before and after the current luminance value can also be regarded as the number of occurrences of adjacent current luminance values, which is not limited in the present invention.

It can be seen from the histogram Histogram in FIG. 3 and the smoothed histogram Histogram1 in FIG. 4 that the image processor 120 can eliminate the statistics caused by noise through smoothing filtering, that is, the luminance values 253, 255, 256 of the histogram. The occurrences H(253), H(255), H(256), H(258) and H(259) of the , 258 and 259 mappings are noise.

Next, after step S233, the image processor 120 searches the smoothed histogram forward from the first effective luminance value for the first luminance value with the number of occurrences as a second effective luminance value, and uses the first effective luminance value as a second effective luminance value. The first luminance value to the second effective luminance value is used as a second range (step S235). Continuing the above example and referring to FIG. 4 , the image processor 120 searches the smoothed histogram Histogram1 from the first effective luminance value L1 forward to search for the first luminance value 252 with the number of occurrences as the second effective luminance value L2 . The image processor 120 then uses the first luminance value L0 to the second effective luminance value L2 as the second range DR2.

After step S235, the image processor 120 will further determine in the second range DR2 whether the second effective luminance value L2 is less than or equal to a predetermined luminance value (step S237). If the second effective brightness value L2 is less than or equal to the preset brightness value, it means that the preset brightness value can cover all the brightness values in the second range DR2. At this time, the image processor 120 uses the first luminance value L0 to the preset luminance value as the effective dynamic range (step S238 ).

Conversely, if the second effective luminance value L2 is greater than the preset luminance value, it means that the preset luminance is insufficient to cover all the luminance values in the second range DR2. At this time, the image processor 120 uses the first luminance value L0 to the second effective luminance value as the effective dynamic range (step S239). It is worth noting that the preset luminance value can be set according to the resolution of the second range DR2, the resolution of the input luminance values P0-Pn, or other luminance values associated with the smoothed histogram Histogram1, which is not implemented in the present invention. limit.

In this embodiment, the preset brightness value is set to 255. Therefore, following the above example, the image processor 120 will determine that the second effective luminance value L2 is less than or equal to 255 in the second range DR2. At this time, the image processor 120 uses the first luminance value L0 to the preset luminance value (ie, 255) as the effective dynamic range Deff. Therefore, the image processor 120 can define an appropriate effective dynamic range through steps S231 to S239 for subsequent processing.

Returning to FIG. 2 again, after determining the effective dynamic range (ie, step S230 ), the image processor 120 sequentially accumulates each occurrence of smoothing in the effective dynamic range to generate an accumulation curve. The accumulation curve will represent the relationship between these brightness values and the accumulated occurrence times (step S240). Following the above example, the image processor 120 sequentially accumulates the number of occurrences of smoothing H'(0)-H'(255) corresponding to each luminance value 0-255 in the effective dynamic range Deff shown in FIG. 4 , to produce the accumulated number of occurrences Had(n). Therefore, the accumulated number of occurrences Had(0)=13, Had(1)=13+12=25, and Had(2)=13+12+12=37. The calculation method of other accumulated occurrence times Had(3)-Had(255) is roughly the same as that of Had(1), and the calculation result is shown in the accumulated curve Cuv in Figure 5, so it is not repeated here. .

After step S240, the image processor 120 normalizes the accumulated occurrence times to the effective dynamic range to generate an output luminance value, and the luminance values and the output luminance value corresponding to each luminance value form a global mapping curve (step S240). S250). More specifically, the image processor 120 will sequentially calculate the proportional relationship between the accumulated number of occurrences Had(n) and all the number of occurrences in the effective dynamic range, and multiply each proportional relationship by a value in the effective dynamic range. The highest luminance value to generate the output luminance value Iout(n).

Taking the accumulated number of occurrences Had(0)=13 in FIG. 5 as an example, the total number of occurrences in the effective dynamic range Deff is 147, and the highest luminance value in the effective dynamic range Deff is 255. Therefore, the proportional relationship between the accumulated number of occurrences Had(0) and the total number of occurrences is (13/147). The output luminance value Iout(0) is the proportional relationship multiplied by the highest luminance value=(13/147)*255=23. Taking the accumulated number of occurrences Had(1)=25 in FIG. 5 as an example, the total number of occurrences in the effective dynamic range Deff is 147, and the highest luminance value in the effective dynamic range Deff is 255. Therefore, the proportional relationship between the accumulated number of occurrences Had(1) and the total number of occurrences is (25/147). The output luminance value Iout(1) is the proportional relationship multiplied by the highest luminance value=(25/147)*255=43.

The calculation method of other output luminance values Iout(2)-Iout(255) is roughly the same as that of Iout(1), and the calculation result is shown in the cumulative curve Cuv of FIG. Accordingly, these luminance values 0-255 and the output luminance values Iout(0)-Iout(255) corresponding to each luminance value 0-255 will form the global mapping curve Cgb.

After obtaining the global mapping curve, the image processor 120 sequentially obtains the corresponding output luminance value according to the input luminance value of each pixel position (step S260 ). For example, please refer to FIG. 6 to FIG. 7 at the same time, the input image Fr1 has 10*15 pixel positions P0-P149, and each pixel position P0-P149 has an input luminance value, such as the input luminance value of the pixel position P22 3. The input luminance value of pixel position P47 is 250, and the input luminance value of pixel position P102 is 4.

Therefore, the image processor 120 corresponds the input luminance value 3 of the pixel position P22 to the luminance value 3 in the global mapping curve Cgb, and obtains the output luminance value Iout(3)=85 corresponding to the luminance value 3 . The image processor 120 corresponds the input luminance value 250 at the pixel position P47 to the luminance value 250 in the global mapping curve Cgb, and obtains the output luminance value Iout(250)=243 corresponding to the luminance value 250. The image processor 120 corresponds the input luminance value 4 of the pixel position P102 to the luminance value 4 in the global mapping curve Cgb, and obtains the output luminance value Iout(4)=104 corresponding to the luminance value 4. The input luminance values of other pixel positions can also find the corresponding output luminance values in this way, so they will not be repeated here.

After step 260, the image processor 120 adjusts the corresponding output luminance value at each pixel position according to a luminance relationship between the corresponding input luminance value and a plurality of adjacent input luminance values to generate a final luminance value (step S270). Furthermore, since the input luminance value is composed of incident light (eg, the low-frequency portion of the input luminance value) and reflected light (eg, the high-frequency portion of the input luminance value), and in this embodiment, the input luminance value=incident light* reflected light. If the image processor 120 can remove the low frequency part, it can enhance the high frequency part.

Therefore, please refer to FIG. 2B at the same time, the image processor 120 will calculate at least one high frequency pixel ratio according to the corresponding input luminance value and the adjacent input luminance value, and use the at least one high frequency pixel ratio as the luminance relationship (step S271 ). The at least one high frequency pixel ratio is associated with the input luminance value of the corresponding pixel position and the at least one low frequency pixel value.

In this embodiment, since there are many factors affecting the incident light, the image processor 120 uses at least one low-frequency wave filter to simulate different incident light. Therefore, the image processor 120 will calculate at least one low-frequency pixel value according to the corresponding input luminance value and the adjacent input luminance value, and calculate the proportional relationship between the corresponding input luminance value and the at least one low-frequency pixel value, so as to generate at least one high-frequency pixel Proportion.

For illustration, the pixel position P22 of the input image Fr1 in FIG. 7 and the image processor 120 calculate two low-frequency pixel values through an averaging filter with a 3*3 mask and a 5*5 mask. The input luminance value of the pixel position P22 and its adjacent input luminance values form a pixel group F1 and are shown in FIG. 8 . The image processor 120 will calculate a low frequency pixel value (ie (1+2+3+4+5+1+2+3+4+5+1+2+3+ 4+5+1+2+3+4+5+1+2+3+4+5)/25=3). Next, the image processor 120 will calculate the ratio between the corresponding input luminance value and the low-frequency pixel value to generate a high-frequency pixel ratio (ie, 3/3=1).

Similarly, the image processor 120 will calculate another low frequency pixel value (ie (2+3+4+2+3+4+2+3+4)/9=3 through an averaging filter with a 3*3 mask ). Next, the image processor 120 will calculate the ratio between the corresponding input luminance value and another low-frequency pixel value to generate another high-frequency pixel ratio (ie, 3/3=1). The image processor 120 then multiplies the above two high-frequency pixel ratios (ie, 1*1=1) to obtain the luminance relationship.

After obtaining the luminance relationship (ie, step S271 ), the image processor 120 will adjust the corresponding output luminance value according to the luminance relationship to generate a final luminance value (step S273 ). In this embodiment, the last luminance value=luminance relation*output luminance value=1*3=3, which means that the image processor 120 adjusts the corresponding output luminance value (=3) according to the luminance relation (=1) to generate the final luminance value (=3). When the final luminance value is greater than a maximum luminance value in the effective dynamic range (255 in this embodiment), the image processor 120 takes the highest luminance value as the final luminance value. Of course, the image processor 120 can also adjust the corresponding output brightness value by other calculation methods and brightness relationship to generate the final brightness value, which is not limited in the present invention.

It can be seen from the above that the difference between the input luminance values of the pixel position P22 in the pixel group F1 and these adjacent input luminance values is very small (ie, evenly distributed). Therefore, the image processor 120 does not need to adjust the output luminance value so that the final luminance value of the pixel position P22 is equal to the output luminance value.

Next, the pixel position P47 of the input image Fr1 in FIG. 7 and the image processor 120 calculate two low-frequency pixel values through an averaging filter with a 3*3 mask and a 5*5 mask are used for illustration. The input luminance value of the pixel position P47 and its adjacent input luminance values form a pixel group F2 and are shown in FIG. 9 . One of the low frequency pixel value = (6+6+7+8+8+6+6+250+7+8+6+6+250+10+7+9+9+250+10+10+9+9 +10+10+10)/25=37.08, and the corresponding high frequency pixel ratio=250/37.08=6.74. Another low frequency pixel value=(6+250+7+6+250+10+9+250+10)/9=88.67, and the corresponding high frequency pixel ratio=250/88.67=2.82. And the luminance relation=6.74*2.82=19. Last luminance value=luminance relation*output luminance value=19*250=4752. The final luminance value is greater than a maximum luminance value of 255 in the effective dynamic range Deff, so the image processor 120 uses the highest luminance value of 255 as the final luminance value.

It can be seen from the above that the input luminance value of the pixel position P47 in the pixel group F2 is very different from these adjacent input luminance values, and the input luminance value of the pixel position P47 is higher than these adjacent input luminance values. Therefore, the image processor 120 will brighten the output luminance value, so that the difference between the final luminance value of the pixel position P47 and the adjacent input luminance value is larger, so as to further improve the contrast ratio of the pixel position P47.

Next, the pixel position P102 of the input image Fr1 in FIG. 7 and the image processor 120 calculate two low-frequency pixel values through an averaging filter with a 3*3 mask and a 5*5 mask for illustration. The input luminance value of the pixel position P102 and its adjacent input luminance values form a pixel group F3 and are shown in FIG. 10 . One of the low frequency pixel value = (249+249+249+249+249+249+249+249+249+249+1+4+4+4+1+250+250+250+250+250+251+251 +251+253+255)/25=200.6, and the corresponding high frequency pixel ratio=4/200.6=0.02. Another low frequency pixel value=(249+249+249+4+4+4+250+250+250)/9=167.7, and the corresponding high frequency pixel ratio=4/167.7=0.02. And the luminance relation=0.02*0.02=0. Last luminance value=luminance relation*output luminance value=0*4=0.

It can be seen from the above that the input luminance value of the pixel position P102 in the pixel group F3 is very different from these adjacent input luminance values, and the input luminance value of the pixel position P102 is lower than these adjacent input luminance values. Therefore, the image processor 120 dims the output luminance value, so that the difference between the final luminance value of the pixel position P102 and the adjacent input luminance value is larger, so as to further improve the contrast ratio of the pixel position P102.

Therefore, from the pixel groups F1 to F3 of the above input image Fr1, it can be known that when the difference between the input luminance values in the pixel group is small (such as the pixel group F1), it means that the current pixel position (such as the pixel position P22) is not in the input image Fr1. In the edge portion, the image processor 120 does not adjust the output luminance value of the current pixel position, or slightly adjusts the output luminance value of the current pixel position according to the difference value.

When the difference between the input luminance values in the pixel group is very large (such as the pixel groups F2 and F3), it means that the current pixel position (such as the pixel positions P47 and 102) is the edge part of the input image Fr1, and the image processor 120 will The value of the gap, the value of the current input brightness and the value of the adjacent input brightness value to adjust the output brightness value of the current pixel position. As mentioned above, the input luminance value of the pixel position P47 is very different from the adjacent input luminance values, and the input luminance value of the pixel position P47 is higher than these adjacent input luminances. For another example, the input luminance value of the above-mentioned pixel position P102 is very different from the adjacent input luminance values, and the input luminance value of the pixel position P102 is lower than these adjacent input luminances.

Therefore, the image processor 120 can adaptively adjust the output luminance value of the current pixel position according to the luminance relationship between the input luminance value of the current pixel position and its adjacent input luminance values, so as to generate the final luminance value accordingly.

To sum up, the embodiments of the present invention provide a contrast enhancement method based on dynamic range compression and an electronic device thereof, which determine an appropriate effective dynamic range by the number of occurrences of the input luminance value of each pixel position in an input image. Then, according to the regional characteristics of each pixel position in the input image, the output luminance value of the input luminance value mapping is adjusted to adaptively improve the contrast of the input image. Accordingly, the contrast enhancement method and the electronic device thereof of the present invention can reduce computational complexity and generate better image contrast.

The above descriptions are merely embodiments of the present invention, and are not intended to limit the claims of the present invention.

Claims (10)

1. A contrast enhancement method based on dynamic range compression is applicable to an electronic device and used for adjusting an input brightness value of each pixel position in an input image so as to enhance the contrast of the input image, and the contrast enhancement method comprises the following steps:

receiving each input brightness value in the input image;

corresponding the occurrence frequency of each input brightness value to a plurality of brightness values on a histogram, carrying out smooth filtering on the occurrence frequency, and determining an effective dynamic range according to the smoothed occurrence frequency;

Sequentially accumulating each of the smoothed occurrences in the effective dynamic range to generate an accumulation curve, wherein the accumulation curve represents a relationship between the luminance value and the accumulated occurrences;

normalizing the accumulated occurrence number to the effective dynamic range to generate an output brightness value, wherein the brightness value and the output brightness value corresponding to each brightness value form a global mapping curve;

in the global mapping curve, obtaining the corresponding output brightness value according to the input brightness value of each pixel position in sequence; and

in each of the pixel positions, the corresponding output luminance value is adjusted according to a luminance relationship between the corresponding input luminance value and a plurality of neighboring input luminance values to generate a final luminance value.

2. The contrast enhancement method based on dynamic range compression as claimed in claim 1, wherein the step of determining the effective dynamic range further comprises:

searching a first brightness value with occurrence times from the last brightness value in the histogram as a first effective brightness value, and taking the first brightness value to the first effective brightness value as a first range;

In the first range, smoothing and filtering the corresponding occurrence number to generate a smoothed histogram;

searching a first brightness value with occurrence times from the first effective brightness value to the second effective brightness value in the smoothed histogram as a second effective brightness value, and taking the first brightness value to the second effective brightness value as a second range; and

in the second range, determining whether the second effective brightness value is less than or equal to a predetermined brightness value, if the second effective brightness value is less than or equal to the predetermined brightness value, using the first brightness value to the predetermined brightness value as the effective dynamic range, and if the second effective brightness value is greater than the predetermined brightness value, using the first brightness value to the second effective brightness value as the effective dynamic range.

3. The dynamic range compression-based contrast enhancement method of claim 1, wherein the step of normalizing the accumulated number of occurrences to the effective dynamic range further comprises:

calculating the proportional relation between the accumulated occurrence times and all the occurrence times in the effective dynamic range in sequence; and

multiplying each of the proportional relations by a highest luminance value in the effective dynamic range to generate the output luminance value.

4. The method of claim 1, wherein the step of adjusting the corresponding output luminance value to generate the final luminance value in each pixel position further comprises:

calculating at least one high-frequency pixel proportion according to the corresponding input brightness value and the adjacent input brightness value, and taking the at least one high-frequency pixel proportion as the brightness relation, wherein the at least one high-frequency pixel proportion is associated with the input brightness value and at least one low-frequency pixel value of the corresponding pixel position; and

and adjusting the corresponding output brightness value according to the brightness relation to generate the final brightness value.

5. The contrast enhancement method according to claim 4, wherein the step of calculating the at least one high frequency pixel proportion corresponding to the pixel position further comprises:

calculating the at least one low-frequency pixel value according to the corresponding input brightness value and the adjacent input brightness value; and

calculating a proportional relationship between the corresponding input luminance value and the at least one low frequency pixel value to generate the at least one high frequency pixel proportion.

6. The contrast enhancement method according to claim 4, wherein the step of adjusting the corresponding output luminance value according to the luminance relationship further comprises:

Multiplying the corresponding output brightness value by the corresponding brightness relation to generate the final brightness value, and taking the maximum brightness value as the final brightness value when the final brightness value is larger than a maximum brightness value in the effective dynamic range.

7. An electronic device based on dynamic range compression for adjusting an input luminance value of each pixel position in an input image to enhance contrast of the input image, the electronic device comprising:

an image acquisition device for receiving the input image and sequentially acquiring each input brightness value in the input image; and

an image processor electrically connected to the image acquisition device for executing the following steps:

receiving each input brightness value in the input image;

corresponding the occurrence frequency of each input brightness value to a plurality of brightness values on a histogram, carrying out smooth filtering on the occurrence frequency, and determining an effective dynamic range according to the smoothed occurrence frequency;

sequentially accumulating each of the smoothed occurrences in the effective dynamic range to generate an accumulation curve, wherein the accumulation curve represents a relationship between the luminance value and the accumulated occurrences;

Normalizing the accumulated occurrence number to the effective dynamic range to generate an output brightness value, wherein the brightness value and the output brightness value corresponding to each brightness value form a global mapping curve;

in the global mapping curve, obtaining the corresponding output brightness value according to the input brightness value of each pixel position in sequence; and

in each of the pixel positions, the corresponding output luminance value is adjusted according to a luminance relationship between the corresponding input luminance value and a plurality of neighboring input luminance values to generate a final luminance value.

8. The electronic device according to claim 7, wherein in normalizing the accumulated occurrences, the image processor sequentially calculates a ratio of the accumulated occurrences to all occurrences in the effective dynamic range, and multiplies a highest luminance value in the effective dynamic range by each ratio to generate the output luminance value.

9. The electronic device of claim 7, wherein when adjusting the corresponding output luminance value to generate the final luminance value in each of the pixel positions, the image processor calculates at least one high frequency pixel ratio according to the corresponding input luminance value and the neighboring input luminance value, uses the at least one high frequency pixel ratio as the luminance relationship, and adjusts the corresponding output luminance value according to the luminance relationship to generate the final luminance value, wherein the at least one high frequency pixel ratio is associated with the input luminance value and at least one low frequency pixel value of the corresponding pixel position.

10. The electronic device of claim 9, wherein the image processor calculates the at least one low-frequency pixel value according to the corresponding input luminance value and the neighboring input luminance value, calculates a ratio of the corresponding input luminance value and the at least one low-frequency pixel value to generate the at least one high-frequency pixel ratio as the luminance relationship, and multiplies the corresponding output luminance value by the luminance relationship to generate the final luminance value, wherein the image processor uses the highest luminance value as the final luminance value when the final luminance value is greater than a highest luminance value in the effective dynamic range.

Images