CN110674782B - Infrared dim target detection method based on fractional entropy - Google Patents
Infrared dim target detection method based on fractional entropy Download PDFInfo
- Publication number
- CN110674782B CN110674782B CN201910947684.9A CN201910947684A CN110674782B CN 110674782 B CN110674782 B CN 110674782B CN 201910947684 A CN201910947684 A CN 201910947684A CN 110674782 B CN110674782 B CN 110674782B
- Authority
- CN
- China
- Prior art keywords
- entropy
- image
- sliding
- fractional
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/20—Scenes; Scene-specific elements in augmented reality scenes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an infrared small and weak target detection method based on fractional entropy, belongs to the field of original infrared image processing and target detection, and solves the problem that the infrared small and weak target in the prior art is difficult to detect. According to the method, a sliding window is established on an obtained original infrared image to be detected, and a standby entropy value image is obtained according to the sliding window, the sliding step length of the sliding window and the original infrared image; based on the sliding step length of the sliding window, sliding the sliding window once on the original infrared image from left to right and from top to bottom to obtain an ROI (region of interest), calculating the fractional entropy of the ROI and assigning the fractional entropy to a pixel corresponding to a standby entropy value image, and continuously sliding the window until the whole original infrared image is traversed; mapping pixel values in the assigned fractional entropy image into a space [0,255] to obtain an entropy value image; and after morphological processing is carried out on the entropy image, determining the position and the size of a target, and carrying out repositioning on the original image to obtain a detection result.
Description
Technical Field
An infrared dim target detection method based on fractional entropy is used for infrared dim target detection and belongs to the field of original infrared image processing and target detection.
Background
With the advancement of technology, imaging weapon systems are receiving more and more national attention and development, and are applied to various aspects, such as: missile early warning, missile interception, fire control systems, post-disaster monitoring systems and the like, and infrared detection is favored in imaging weapon systems due to the characteristic of all-weather work of the infrared detection. Imaging weapon systems usually use infrared remote detection, so the target detected by the imaging weapon systems often has the characteristics of small imaging area, weak radiation energy, blurred target edge, low signal-to-noise ratio and the like, and the accurate detection of the target is difficult. The current infrared weak and small target detection comprises the following steps: detection-before-tracking (DBT), tracking-before-detection (TBD), single frame image detection. The DBT algorithm is used when the signal-to-noise ratio is high, and when the signal-to-noise ratio is low, the detection effect is not ideal. The TBD algorithm has a good detection effect when the signal-to-noise ratio is low, but the algorithm is complex in calculation, needs to perform global traversal for multiple times, and is difficult to apply to hardware in a large quantity. The single-frame image detection method comprises a frequency domain algorithm and a space domain algorithm. The frequency domain method requires the target to have a higher frequency domain and the background to have a lower frequency domain, but the background clutter in the actual site will also have a higher frequency domain, often causing false detection. The airspace method mainly utilizes the gray characteristic to detect the target, but the infrared weak target and the background have similar brightness and cannot be accurately detected.
Disclosure of Invention
Aiming at the problems of the research, the invention aims to provide an infrared dim target detection method based on fractional entropy, which solves the problem of low detection precision in the prior art under the condition of low signal-to-noise ratio and low image contrast.
In order to achieve the purpose, the invention adopts the following technical scheme:
an infrared dim target detection method based on fractional entropy comprises the following steps:
s1, establishing a sliding window on an obtained original infrared image to be detected, and obtaining a standby entropy value image according to the sliding window, a sliding step length of a given sliding window and the original infrared image;
s2, sliding on the original infrared image once from left to right and from top to bottom based on the sliding step length of the sliding window to obtain an ROI (region of interest), calculating the fractional entropy of the ROI and assigning the fractional entropy to a pixel corresponding to the standby entropy value image;
s3, judging whether the sliding window traverses the complete original infrared image, if so, obtaining a fractional entropy image, mapping a pixel value in the fractional entropy image into a space [0,255] to obtain a mapped entropy image, and if not, turning to the step S2 to continue sliding the sliding window;
and S4, performing morphological filtering processing on the mapped entropy image to obtain the position and the size of the target, performing repositioning on the original infrared image based on the position and the size of the target to determine the target, and finally obtaining a detection result.
Further, the specific steps of step S1 are:
s1.1, establishing a sliding window on the obtained original infrared image to be detected, and giving the sliding step length of the sliding window;
s1.2, obtaining the size of the standby entropy image according to the sliding window, the sliding step length of the sliding window and the original infrared image, wherein the specific formula is as follows:
wherein, I 2 Cols and I 2 Rows is a spare entropy value image I 2 The floor () represents a downward integer, N is the pixel column number of the original infrared image, M is the pixel row number of the original infrared image, N is the pixel column number of the sliding window, M is the pixel row number of the sliding window, w is the sliding step length in the horizontal direction, and h is the sliding step length in the vertical direction;
and S1.3, directly creating a spare entropy value image based on the size of the spare entropy value image.
Further, the size of the sliding window is 6 × 6, and the sliding step size in the horizontal direction and the sliding step size in the vertical direction are both 1.
Further, the specific steps of step S2 are:
s2.1, sliding on the original infrared image once based on the sliding step length of the sliding window in a mode of from left to right and from top to bottom to obtain an ROI (region of interest);
s2.2, quantizing the gray value of the pixel in the ROI area based on the quantization with the gradient Md, wherein Md is the gradient value after gray quantization, the distribution condition of the gray value of the ROI area is totally measured by a one-dimensional array alpha [ ] initialized to 0, wherein alpha [0] corresponds to the number of the pixel with the gray value of 0 in the ROI area, alpha [1] corresponds to the number of the pixel with the gray value Md after quantization in the ROI area, the number of the pixel with the gray value k × Md after quantization in the ROI area corresponding to alpha [ k ] is analogized in sequence, finally, the distribution condition of the gray value of the pixel in the ROI area is obtained, and the total number of the pixels is obtained according to the distribution condition:
s2.3, calculating the proportion of each gray value in the alpha [ ] to the total number of the pixels, and correspondingly storing the proportion in an array b [ ], wherein the specific formula is as follows:
s2.4, according to the array b [ ]]Calculating fractional entropy S within the ROI area α The formula is as follows:
wherein, the value range of alpha is 0< alpha <1;
and S2.5, assigning the obtained fractional entropy value to a pixel corresponding to the spare entropy value image.
Further, the specific step of step S2.5 is:
s2.5.1, sliding a sliding window on the original infrared image from left to right and from top to bottom to obtain ROI areas, wherein the ROI areas obtained by sliding the sliding window horizontally correspond to pixels in the standby entropy value image one by one, namely the number of the ROI areas obtained by sliding the sliding window horizontally is the same as the number of the pixels in each line of the standby entropy value image;
and S2.5.2, assigning the fractional entropy to the pixel corresponding to the spare entropy value image based on the corresponding relation.
Further, in step S3, mapping the pixel values in the fractional entropy image into a space [0,255], and obtaining an entropy image specifically includes:
s3.1, comparing all pixel values in the fractional entropy image, obtaining a maximum value Max in the classified entropy image after comparison, and calculating a scaling eta of the maximum value which is mapped to Max:
max is the maximum gray value which can be represented by the fractional entropy image, and for the n-bit original infrared image, the Max value is 2 n -1；
S3.2, based on the scaling eta, calculating the value y of mapping each pixel x in the fractional entropy image to a space [0,255], and obtaining the mapped entropy image, wherein the formula is as follows:
y [0,255] ＝η·x。
further, the specific steps of step S4 are:
s4.1, performing morphological filtering on the entropy image, namely performing corrosion and expansion, removing noise points interfering with target detection, and obtaining the position (x, y) of the upper left corner of the target and the approximate size (width, height);
s4.2, determining a target range on the original infrared image based on the position (x, y) of the upper left corner of the target and the approximate size (width, height):
I 1 .x＝x*w-p
I 1 .y＝y*h-p
I 1 .width＝width*w+2*p
I 1 .height＝heigh*h+2*p
wherein, I 1 X and I 1 Y is the upper left corner position of the object in the original image, I 1 Width and I 1 Height is the original infrared image I 1 The width and the height of the medium target are p is an addition and subtraction coefficient, and the value is 3, so that the range of the target is properly enlarged, the target to be detected is ensured to fall into an expected range, and the target range in the original drawing can be obtained；
S4.3, directly regarding the pixels outside the target range as a background based on the target range, namely performing zeroing processing on the pixels outside the target range, and performing binarization processing on the pixels within the target range to obtain a complete target:
wherein t is a threshold value determined according to actual conditions.
Compared with the prior art, the invention has the beneficial effects that:
1. according to the fractional entropy image calculated according to the original infrared image, the edge and texture information of the target in the image can be well represented, so that the target can be detected when the edge of the target is fuzzy.
2. The method can well distinguish the target from the noise, can accurately detect the target when the noise is large, can directly calculate the fractional entropy of the image without carrying out noise reduction processing on the image, and can avoid target loss and target detail loss caused by image filtering.
3. The method is easy to realize, accurate in detection effect, less in calculation steps and higher in algorithm speed compared with a plurality of algorithms.
Drawings
FIG. 1 is a flow chart of a method of the present invention;
FIG. 2 is an original infrared image of an infrared weak and small target with blurred edge and low contrast;
FIG. 3 is a fractional entropy image obtained based on FIG. 2 in accordance with the present invention;
FIG. 4 is a graph of the detection results obtained on the original IR image of FIG. 2 in accordance with the present invention;
FIG. 5 is a schematic diagram of a sliding window established on an original infrared image according to the present invention;
FIG. 6 is an original infrared image of an infrared weak small target with a large background noise;
FIG. 7 is a fractional entropy image obtained based on FIG. 6 in the present invention;
fig. 8 is a graph of the detection results obtained on the raw infrared image of fig. 6 in accordance with the present invention.
Detailed Description
The invention will be further described with reference to the accompanying drawings and the detailed description.
An infrared dim target detection method based on fractional entropy comprises the following steps:
s1, establishing a sliding window on an obtained original infrared image to be detected, and obtaining a standby entropy value image according to the sliding window, a sliding step length of a given sliding window and the original infrared image;
the method comprises the following specific steps:
s1.1, establishing a sliding window on the obtained original infrared image to be detected, giving a sliding step length of the sliding window, wherein the size of the sliding window is 6x6, the sliding step lengths in the horizontal direction and the vertical direction are both 1, and the size of the sliding window and the sliding step length can also take other values;
s1.2, obtaining the size of the standby entropy image according to the sliding window, the sliding step length of the sliding window and the original infrared image, wherein the specific formula is as follows:
wherein, I 2 Cols and I 2 Rows is a spare entropy value image I 2 The floor () represents a downward integer, N is the pixel column number of the original infrared image, M is the pixel row number of the original infrared image, N is the pixel column number of the sliding window, M is the pixel row number of the sliding window, w is the sliding step length in the horizontal direction, and h is the sliding step length in the vertical direction;
s1.3, directly creating a standby entropy value image based on the size of the standby entropy value image;
s2, sliding on the original infrared image once from left to right and from top to bottom based on the sliding step length of the sliding window to obtain an ROI (region of interest), calculating the fractional entropy of the ROI and assigning the fractional entropy to a pixel corresponding to the spare entropy value image;
the method comprises the following specific steps:
s2.1, sliding the original infrared image once on the basis of the sliding step length of the sliding window in a mode of going from left to right and going from top to bottom to obtain an ROI (region of interest);
s2.2, based on the quantization with the gradient Md, md is the gradient value after the gray scale quantization, the pixel gray value in the ROI area is quantized, the distribution condition of the gray scale value of the ROI area after the one-dimensional array alpha [ ] which is initialized to 0 is used for carrying out the statistical measurement, wherein alpha 0 corresponds to the pixel number with the gray scale value of 0 in the ROI area, alpha 1 corresponds to the pixel number with the gray scale value of Md after the quantization in the ROI area, and the pixel number with the gray scale value of k Md after the alpha k corresponds to the quantization in the ROI area is analogized in sequence, and finally the pixel gray scale value distribution condition in the ROI area is obtained. Such as: based on the quantization with 4 gradient, quantizing the pixel gray value in ROI area, using the one-dimensional array alpha [ ] initialized to 0 to count the distribution of gray value in ROI area, where alpha 0 corresponds to the pixel number with 0 gray value in ROI area, alpha 1 corresponds to the pixel number with 4 gray value in ROI area, and so on (for example, alpha 2 corresponds to the pixel number with 8 gray value after quantization in ROI area, alpha 3 corresponds to the pixel number with 12 gray value after quantization in ROI area, and alpha 4 corresponds to the pixel number with 16 gray value after quantization in ROI area), finally obtaining the pixel gray value distribution in ROI area, and obtaining the total number of pixels according to the distribution:
the quantization with the gradient of 4 is selected to be the best in the scheme, and the larger the quantization value is, the faster the fractional entropy calculation is, but the overall accuracy is reduced;
s2.3, calculating the proportion of each gray value in the alpha [ ] to the total number of the pixels, and correspondingly storing the proportion in an array b [ ], wherein the specific formula is as follows:
s2.4, according to array b]Calculating fractional entropy S within the ROI region α The formula is as follows:
wherein, the value range of alpha is 0< alpha <1;
and S2.5, assigning the obtained fractional entropy value to a pixel corresponding to the spare entropy value image.
The method comprises the following specific steps:
s2.5.1, sliding a sliding window on the original infrared image from left to right and from top to bottom to obtain ROI areas, wherein the ROI areas obtained by sliding the sliding window horizontally correspond to pixels in the standby entropy value image one by one, namely the number of the ROI areas obtained by sliding the sliding window horizontally is the same as the number of the pixels in each line of the standby entropy value image;
and S2.5.2, assigning the fractional entropy to the pixel corresponding to the spare entropy value image based on the corresponding relation.
S3, judging whether the sliding window traverses the original infrared image, if so, mapping the pixel value in the fractional entropy image into a space [0,255] to obtain a mapped entropy image, and if not, turning to the step S2 to continue sliding the window;
mapping pixel values in the fractional entropy image into a space [0,255], wherein the specific steps of obtaining the entropy image are as follows:
s3.1, comparing all pixel values in the fractional entropy image, obtaining a maximum value Max in the classified entropy image after comparison, and calculating a scaling eta of the maximum value which is mapped to Max:
wherein Max is the maximum that the fractional entropy image can be characterizedGray value, max value is 2 for n-bit original infrared image n -1；
S3.2, based on the scaling eta, calculating the value y of mapping each pixel x in the fractional entropy image to a space [0,255], and obtaining the mapped entropy image, wherein the formula is as follows:
y [0,255] ＝η·x。
and S4, performing morphological filtering processing on the mapped entropy image to obtain the position and the size of the target, performing repositioning on the original infrared image to determine the target, and finally obtaining a detection result.
The method comprises the following specific steps:
s4.1, performing morphological filtering on the entropy image, namely, firstly performing corrosion and then performing expansion, removing noise points which interfere with target detection, and obtaining the position (x, y) of the upper left corner and the approximate size (width, height) of the target;
s4.2, determining a target range on the original infrared image based on the position (x, y) of the upper left corner of the target and the approximate size (width, height):
I 1 .x＝x*w-p
I 1 .y＝y*h-p
I 1 .width＝width*w+2*p
I 1 .height＝heigh*h+2*p
wherein, I 1 X and I 1 Y is the upper left corner position of the target in the original image, I 1 Width and I 1 Height is the original infrared image I 1 The width and the height of the medium target are p is an addition and subtraction coefficient, and the value is 3, so that the range of the target is properly enlarged, the target to be detected is ensured to fall into an expected range, and the target range in the original drawing can be obtained;
s4.3, directly regarding the pixels outside the target range as a background based on the target range, namely, performing zeroing processing on the pixels outside the target range, and performing binarization processing on the pixels within the target range to obtain a complete target:
where t is a threshold determined according to actual conditions, the threshold is given as 50 in the following embodiments.
Example 1
Establishing a 6x6 sliding window on the obtained original infrared image with the pixel to be detected being 320x240, and obtaining a 315x235 standby entropy value image according to the 6x6 sliding window, the sliding step length of the given sliding window being 1x1 and the original infrared image, wherein the original infrared image is shown in fig. 2 and has a fuzzy edge;
based on the sliding step length of the sliding window, sliding the window on the original infrared image from left to right and from top to bottom, obtaining an ROI (region of interest) by sliding each time, calculating the fractional entropy of the ROI and assigning the fractional entropy to a pixel corresponding to the standby entropy value image to obtain a fractional entropy image;
after the sliding window finishes sliding the original infrared image, mapping the pixel values in the fractional entropy image into a space [0,255] to obtain a mapped entropy image, as shown in fig. 3;
performing morphological filtering on the entropy image, performing corrosion and expansion firstly, removing noise points interfering with target detection, and obtaining the position (x, y) and the approximate size (width, height) of the upper left corner of the target, wherein the coordinates of the upper left corner of the target obtained based on the graph in FIG. 3 are (211, 158), and the width and the height are 13 and 12;
in the original infrared image, the upper left corner position of the target range is (208, 155), and the width and height are 19 and 18;
based on the target range, the pixels outside the target range are directly regarded as the background, that is, the pixels outside the target range are subjected to zeroing processing, and the pixels within the target range are subjected to binarization processing, so that a complete target is obtained, which is shown in fig. 4.
Example 2
Establishing a 6x6 sliding window on the obtained original infrared image with the pixel to be detected being 256 x 172, and obtaining a 251 x 167 spare entropy value image according to the 6x6 sliding window, the sliding step of the given sliding window being 1x1 and the original infrared image, wherein the original infrared image is shown in fig. 6, and the background noise is very high;
based on the sliding step length of the sliding window, sliding the window on the original infrared image from left to right and from top to bottom, obtaining an ROI (region of interest) by sliding each time, calculating the fractional entropy of the ROI and assigning the fractional entropy to a pixel corresponding to the standby entropy value image to obtain a fractional entropy image;
after the sliding window slides the original infrared image, mapping the pixel values in the fractional entropy image into a space [0,255] to obtain a mapped entropy image, as shown in fig. 7;
performing morphological filtering on the entropy image, performing corrosion and expansion firstly, removing noise points interfering with target detection, and obtaining the position (x, y) and the approximate size (width, height) of the upper left corner of the target, wherein the coordinates of the upper left corner of the target obtained based on the graph in FIG. 7 are (152, 105), and the width and the height are 11 and 8;
in the original infrared image, the upper left corner position of the target range is (149, 102), and the width and height are 17 and 14;
based on the target range, the pixels outside the target range are directly regarded as the background, that is, the pixels outside the target range are subjected to zeroing processing, and the pixels within the target range are subjected to binarization processing, so that a complete target is obtained, which is shown in fig. 8.
The above are merely representative examples of the many specific applications of the present invention, and do not limit the scope of the invention in any way. All the technical solutions formed by using the conversion or the equivalent substitution fall within the protection scope of the present invention.
Claims (4)
1. An infrared dim target detection method based on fractional entropy is characterized by comprising the following steps:
s1, establishing a sliding window on an obtained original infrared image to be detected, and obtaining a standby entropy value image according to the sliding window, a sliding step length of a given sliding window and the original infrared image;
s2, sliding on the original infrared image once from left to right and from top to bottom based on the sliding step length of the sliding window to obtain an ROI (region of interest), calculating the fractional entropy of the ROI and assigning the fractional entropy to a pixel corresponding to the standby entropy value image;
s3, judging whether the sliding window traverses the complete original infrared image, if so, obtaining a fractional entropy image, mapping a pixel value in the fractional entropy image into a space [0,255] to obtain a mapped entropy image, and if not, turning to the step S2 to continue sliding the sliding window;
s4, performing morphological filtering processing on the mapped entropy image to obtain the position and the size of a target, performing repositioning on the original infrared image based on the position and the size of the target to determine the target, and finally obtaining a detection result;
the specific steps of the step S1 are as follows:
s1.1, establishing a sliding window on the obtained original infrared image to be detected, and giving a sliding step length of the sliding window;
s1.2, obtaining the size of the backup entropy image according to the sliding window, the sliding step length of the sliding window and the original infrared image, wherein the specific formula is as follows:
wherein, I 2 Cols and I 2 Rows is a spare entropy value image I 2 The floor () represents a downward integer, N is the pixel column number of the original infrared image, M is the pixel row number of the original infrared image, N is the pixel column number of the sliding window, M is the pixel row number of the sliding window, w is the sliding step length in the horizontal direction, and h is the sliding step length in the vertical direction;
s1.3, directly creating a standby entropy value image based on the size of the standby entropy value image;
the specific steps of the step S2 are as follows:
s2.1, sliding the original infrared image once on the basis of the sliding step length of the sliding window in a mode of going from left to right and going from top to bottom to obtain an ROI (region of interest);
s2.2, quantizing the gray value of the pixel in the ROI area based on the quantization with the gradient Md, wherein Md is the gradient value after gray quantization, the distribution condition of the gray value of the ROI area is totally measured by a one-dimensional array alpha [ ] initialized to 0, wherein alpha [0] corresponds to the number of the pixel with the gray value of 0 in the ROI area, alpha [1] corresponds to the number of the pixel with the gray value Md after quantization in the ROI area, the number of the pixel with the gray value k × Md after quantization in the ROI area corresponding to alpha [ k ] is analogized in sequence, finally, the distribution condition of the gray value of the pixel in the ROI area is obtained, and the total number of the pixels is obtained according to the distribution condition:
s2.3, calculating the proportion of each gray value in the alpha [ ] to the total number of the pixels, and correspondingly storing the proportion in an array b [ ], wherein the specific formula is as follows:
s2.4, according to the array b [ ]]Calculating fractional entropy S within the ROI region α The formula is as follows:
wherein, the value range of alpha is 0< alpha <1;
s2.5, assigning the obtained fraction entropy value to a pixel corresponding to the spare entropy value image;
the specific steps of step S2.5 are:
s2.5.1, sliding a sliding window on the original infrared image from left to right and from top to bottom to obtain ROI areas, wherein the ROI areas obtained by sliding the sliding window horizontally correspond to pixels in the spare entropy value image one by one, namely the number of the ROI areas obtained by sliding the sliding window horizontally is the same as the number of the pixels in each line of the spare entropy value image;
and S2.5.2, assigning the fractional entropy to the pixel corresponding to the spare entropy value image based on the corresponding relation.
2. The infrared small target detection method based on fractional entropy of claim 1, wherein the size of the sliding window is 6x6, and the sliding step size in both horizontal and vertical directions is 1.
3. The infrared weak and small target detection method based on fractional entropy as claimed in claim 1, wherein in step S3, mapping pixel values in the fractional entropy image into a space [0,255], and obtaining the entropy image comprises the specific steps of:
s3.1, comparing all pixel values in the fractional entropy image, obtaining a maximum value Max in the classified entropy image after comparison, and calculating a scaling eta of the maximum value which is mapped to Max:
max is the maximum gray value which can be represented by the fractional entropy image, and for the n-bit original infrared image, the Max value is 2 n -1；
S3.2, based on the scaling eta, calculating the value y of each pixel x in the fractional entropy image mapped to a space [0,255] to obtain a mapped entropy image, wherein the formula is as follows:
y [0,255] ＝η·x。
4. the infrared dim target detection method based on fractional entropy as claimed in claim 1, characterized in that the specific steps of step S4 are:
s4.1, performing morphological filtering on the entropy image, namely, firstly performing corrosion and then performing expansion, removing noise points which interfere with target detection, and obtaining the position (x, y) of the upper left corner and the approximate size (width, height) of the target;
s4.2, determining a target range on the original infrared image based on the position (x, y) of the upper left corner of the target and the approximate size (width, height):
I 1 .x＝x*w-p
I 1 .y＝y*h-p
I 1 .width＝width*w+2*p
I 1 .height＝heigh*h+2*p
wherein, I 1 X and I 1 Y is the upper left corner position of the object in the original image, I 1 Width and I 1 Height is the original infrared image I 1 The width and the height of the medium target are p is an addition and subtraction coefficient, and the value is 3, so that the range of the target is properly enlarged, the target to be detected is ensured to fall into an expected range, and the target range in the original drawing can be obtained;
s4.3, directly regarding the pixels outside the target range as a background based on the target range, namely, performing zeroing processing on the pixels outside the target range, and performing binarization processing on the pixels within the target range to obtain a complete target:
wherein t is a threshold value determined according to actual conditions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910947684.9A CN110674782B (en) | 2019-09-30 | 2019-09-30 | Infrared dim target detection method based on fractional entropy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910947684.9A CN110674782B (en) | 2019-09-30 | 2019-09-30 | Infrared dim target detection method based on fractional entropy |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110674782A CN110674782A (en) | 2020-01-10 |
| CN110674782B true CN110674782B (en) | 2022-12-09 |
Family
ID=69080878
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910947684.9A Active CN110674782B (en) | 2019-09-30 | 2019-09-30 | Infrared dim target detection method based on fractional entropy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110674782B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117237619B (en) * | 2023-11-16 | 2024-02-02 | 数聚(山东)医疗科技有限公司 | Water rescue detection system and method based on machine vision technology |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109858483A (en) * | 2019-01-22 | 2019-06-07 | 南京邮电大学 | Small target detecting method based on symbiosis filter form He local entropy |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101051716B1 (en) * | 2010-11-12 | 2011-07-26 | 삼성탈레스 주식회사 | Method for multi sensor image fusion |
| CN105469090B (en) * | 2015-11-19 | 2018-12-18 | 南京航空航天大学 | Small target detecting method and device in infrared image based on frequency-domain residual |
| CN109272489B (en) * | 2018-08-21 | 2022-03-29 | 西安电子科技大学 | Infrared weak and small target detection method based on background suppression and multi-scale local entropy |
| CN109325446B (en) * | 2018-09-19 | 2021-06-22 | 电子科技大学 | Infrared weak and small target detection method based on weighted truncation nuclear norm |
-
2019
- 2019-09-30 CN CN201910947684.9A patent/CN110674782B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109858483A (en) * | 2019-01-22 | 2019-06-07 | 南京邮电大学 | Small target detecting method based on symbiosis filter form He local entropy |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110674782A (en) | 2020-01-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109816641B (en) | Multi-scale morphological fusion-based weighted local entropy infrared small target detection method | |
| CN105787901B (en) | A kind of multiple dimensioned velocity field measurement method for adjacent two interframe in sun full resolution pricture sequence | |
| CN108765335B (en) | Forest fire detection method based on remote sensing image | |
| CN108171193B (en) | Polarized SAR (synthetic aperture radar) ship target detection method based on super-pixel local information measurement | |
| CN102609904A (en) | Bivariate nonlocal average filtering de-noising method for X-ray image | |
| CN103530896A (en) | Image compression and detail enhancement method for infrared image | |
| CN108038856B (en) | Infrared small target detection method based on improved multi-scale fractal enhancement | |
| CN110889843A (en) | SAR image ship target detection method based on maximum stable extremal region | |
| Fan et al. | Dim small target detection based on high-order cumulant of motion estimation | |
| CN108305265B (en) | Real-time processing method and system for weak and small target image | |
| CN110674782B (en) | Infrared dim target detection method based on fractional entropy | |
| CN115063689A (en) | CFAR (computational fluid dynamics) ship target detection method based on local saliency enhancement | |
| CN109886980B (en) | Infrared image cirrus cloud detection method based on neighborhood intensity texture coding | |
| CN107133937B (en) | A kind of self-adapting enhancement method of infrared image | |
| CN109858483B (en) | Small target detection method based on symbiotic filtering morphology and local entropy | |
| CN116740579B (en) | Intelligent collection method for territorial space planning data | |
| CN111243240B (en) | Landslide early warning method and device | |
| CN104268831B (en) | Infrared image compensation method under marine sunlight bright band interference | |
| CN104820972B (en) | A kind of infrared image ME noise remove methods based on in-orbit statistic of classification | |
| CN114926524B (en) | Method for improving measuring accuracy of effective shielding area of infrared smoke screen | |
| CN115980697A (en) | Method for inverting boundary layer height by using laser radar under different weather conditions | |
| CN109461164A (en) | A kind of infrared small target detection method based on direction nuclear reconstitution | |
| CN112598606B (en) | Local self-adaptive infrared image enhancement method based on image decomposition | |
| Ghanbari et al. | Generalized minimum-error thresholding for unsupervised change detection from multilook polarimetric SAR data | |
| Tian et al. | Joint spatio-temporal features and sea background prior for infrared dim and small target detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |