CN109448014B - Image information thinning method based on subgraph - Google Patents

Abstract

The invention relates to an image information thinning method based on a subgraph. The method comprises the following steps: step 1, inputting a gray level image; step 2, calculating a gray level histogram; step 3, calculating a global gray ratio; step 4, calculating a subgraph segmentation threshold; step 5, trimming the gray level histogram; step 6, subgraph segmentation; step 7, calculating a density function; step 8, calculating a distribution function; step 9, constructing a histogram transmission function; step 10, equalization processing; and step 11, fusing subgraphs and outputting. The method can refine local information of the image, and can be applied to the fields of fuzzy digital image enhancement processing and the like.

Description

Image information thinning method based on subgraph

Technical Field

The invention relates to the field of digital image processing, in particular to an image information thinning method based on subgraphs.

Background

Many of the visual effects of image capture are not good due to the influence of scene conditions, which requires image thinning techniques to improve human visual effects, such as highlighting certain features of target objects in the images, extracting characteristic parameters of the target objects from the digital images, and the like, which are all beneficial to the recognition, tracking and understanding of the targets in the images. The main content of the image thinning processing is to highlight interested parts in the image and weaken or remove unnecessary information. This enhances the useful information to obtain a more practical image or to convert it to an image more suitable for human or machine analysis. Currently, commonly used correlation methods include gray level correction, gray level transformation, histogram modification, image smoothing and sharpening, however, the algorithm complexity and the refinement effect still need to be improved.

Disclosure of Invention

The invention provides an image information thinning method based on subgraphs, which adopts the principle of segmenting the subgraphs and adopts methods such as a histogram, a density function, a distribution function and the like respectively, so that the calculation amount is small, and the image thinning effect is good.

The technical scheme adopted for realizing the aim of the invention is as follows: the method comprises the following steps:

step 1: inputting a gray level image I;

step 2: calculating a gray histogram H of the gray image I;

and step 3: calculating the global gray ratio Lg of the gray image;

and 4, step 4: calculating a subgraph segmentation threshold Tseg, wherein Tseg is G (1-Lg);

and 5: in order to prevent the information of the grayscale image I from being lost due to excessive information refinement, the grayscale histogram H needs to be clipped, and when H (I) is greater than the clipping threshold Tc, H (I) ═ Tc, the clipped grayscale histogram is denoted as H_c；

Step 6: segmenting the grayscale image I into I by using a sub-graph segmentation threshold Tseg_LAnd I_HTwo subgraphs, wherein subgraph I_LHas a gray scale value range of [0, Tseg]Scheme I_HThe gray scale value range of (1) is [ Tseg +1, G-1 ]]；

And 7: separately compute subgraph I_LAnd scheme I_HDensity function P of_LAnd P_HWherein P is_L＝H_c/M_L，P_H＝H_c/M_H，M_LAnd M_HAre respectively sub-diagram I_LAnd scheme I_HThe number of pixels of (1);

and 8: separately compute subgraph I_LAnd scheme I_HCumulative distribution function C of_LAnd C_H；

And step 9: respectively using the cumulative distribution function C in step 8_LAnd C_HStructural drawing I_LAnd scheme I_HHas a histogram transfer function TF_LAnd TF_H；

Step 10: for the histogram transfer function TF in step 9 respectively_LAnd TF_HCarrying out equalization processing to respectively obtain equalized subgraphs I_LQAnd I_HQ；

Step 11: merging equalized subgraph I in step 10_LQAnd I_HQObtaining an information refined image I_QAnd output.

The global gray scale ratio Lg in the step 3 is realized by the following formula (1):

in the formula (1), h (I) is the number of pixels with a gray scale value I in the gray scale image I, and G is the gray scale level of the gray scale image I.

The pruning threshold Tc in the step 5 is calculated by the formula (2);

separately computing sub-graph I as described in step 8_LAnd scheme I_HCumulative distribution function C of_LAnd C_HCalculating by the formula (3);

cumulative distribution function C described in step 9_LAnd C_HCumulative distribution function C_LAnd C_HRespectively calculated by the formula (4),

the invention has the beneficial effects that: the method can refine local information of the image, and can be applied to the fields of fuzzy digital image enhancement processing and the like.

Drawings

FIG. 1 is an overall process flow diagram of the present invention.

Detailed Description

The following detailed description of embodiments of the invention refers to the accompanying drawings.

In fig. 1, 101 is a grayscale image input step, 102 is a grayscale histogram calculation step, 103 is a global grayscale ratio calculation step, 104 is a subgraph segmentation threshold calculation step, 105 is a gray top histogram trimming step, 106 is a subgraph segmentation step, 107 is a density function calculation step, 108 is a distribution function calculation step, 109 is a histogram transfer function construction step, 110 is an equalization processing step, and 111 is a subgraph fusion output step.

The specific embodiment process is as follows:

step 101: inputting a gray level image I;

step 102: calculating a gray histogram H of the gray image I;

step 103: calculating the global gray ratio Lg of the image:

in the formula, h (I) is the number of pixels with a gray value I in the gray image I, and G is the gray level of the gray image I;

step 104: calculating a subgraph segmentation threshold Tseg, wherein Tseg is G (1-Lg);

step 105: in order to prevent the information of the grayscale image I from being lost due to excessive information refinement, the grayscale histogram H needs to be clipped, and when H (I) is greater than the clipping threshold Tc, H (I) ═ Tc, the clipped grayscale histogram is denoted as H_cWherein the pruning threshold Tc is calculated by the following formula;

step 106: segmenting the grayscale image I into I by using a sub-graph segmentation threshold Tseg_LAnd I_HTwo subgraphs, wherein subgraph I_LHas a gray scale value range of [0, Tseg]Scheme I_HThe gray scale value range of (1) is [ Tseg +1, G-1 ]]；

Step 107: separately compute subgraph I_LAnd scheme I_HDensity function ofP_LAnd P_HWherein P is_L＝H_c/M_L，P_H＝H_c/M_H，M_LAnd M_HAre respectively sub-diagram I_LAnd scheme I_HThe number of pixels of (1);

step 108: separately compute subgraph I_LAnd scheme I_HDistribution function C of_LAnd C_H；

Step 109: respectively using the distribution function C in step 108_LAnd C_HStructural drawing I_LAnd scheme I_HHas a histogram transfer function TF_LAnd TF_H；

TF_L＝Tseg×C_L

TF_H＝(Tseg+1)+(G-Tseg+1)×C_H

Step 110: respectively for the histogram transfer function TF in step 109_LAnd TF_HCarrying out equalization processing to respectively obtain equalized subgraphs I_LQAnd I_HQ；

Step 111: equalized subgraph I in the fusion step 110_LQAnd I_HQObtaining an information refined image I_QAnd output.

Claims (6)

1. A method for refining image information based on subgraph is characterized by comprising the following steps:

step 1: inputting a gray level image I;

step 2: calculating a gray histogram H of the gray image I;

and step 3: calculating the global gray ratio Lg of the gray image;

and 4, step 4: calculating a subgraph segmentation threshold Tseg, wherein Tseg is G (1-Lg);

and 5: clipping the gray histogram H, and when H (i) is greater than a clipping threshold Tc, H (i) ═ Tc, wherein H (i) is the height of the histogram with the gray value i, and the clipped gray histogram is recorded as H_c；

Step 6: segmenting the grayscale image I into I by using a sub-graph segmentation threshold Tseg_LAnd I_HTwo subgraphs, wherein subgraph I_LHas a gray scale value range of [0, Tseg]Scheme I_HThe gray scale value range of (1) is [ Tseg +1, G-1 ]]；

And 7: separately compute subgraph I_LAnd scheme I_HDensity function P of_LAnd P_H；

And 8: separately compute subgraph I_LAnd scheme I_HDistribution function C of_LAnd C_H；

And step 9: respectively using the distribution function C in step 8_LAnd C_HStructural drawing I_LAnd scheme I_HHas a histogram transfer function TF_LAnd TF_H；

Step 10: for the histogram transfer function TF in step 9 respectively_LAnd TF_HCarrying out equalization processing to respectively obtain equalized subgraphs I_LQAnd I_HQ；

Step 11: merging equalized subgraph I in step 10_LQAnd I_HQObtaining an information refined image I_QAnd output.

2. A sub-picture based image information refining method as claimed in claim 1, wherein said global gray scale value Lg of step 3 is calculated by equation (1):

in the formula (1), h (I) is the number of pixels with a gray scale value I in the gray scale image I, and G is the gray scale level of the gray scale image I.

3. A sub-picture based image information refining method as claimed in claim 1, wherein said pruning threshold Tc of step 5 is calculated by equation (2):

4. a sub-picture based image information refinement method according to claim 1, characterized in that the density function P of step 7_LAnd P_H，P_L＝H_c/M_L，P_H＝H_c/M_H，M_LAnd M_HAre respectively sub-diagram I_LAnd scheme I_HThe number of pixels.

5. A sub-picture based image information refinement method according to claim 1, characterized in that the distribution function C of step 8_LAnd C_HCalculated by equation (3):

6. a sub-picture based image information refinement method according to claim 1, characterized in that the histogram transfer function TF of step 9_LAnd TF_HCalculated by equation (4):

Images