CN101114340A - VLSI realizing system and method of histogram equalization image processing - Google Patents
VLSI realizing system and method of histogram equalization image processing Download PDFInfo
- Publication number
- CN101114340A CN101114340A CNA2007101214807A CN200710121480A CN101114340A CN 101114340 A CN101114340 A CN 101114340A CN A2007101214807 A CNA2007101214807 A CN A2007101214807A CN 200710121480 A CN200710121480 A CN 200710121480A CN 101114340 A CN101114340 A CN 101114340A
- Authority
- CN
- China
- Prior art keywords
- image
- gray
- histogram
- processing
- block
- 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.)
- Pending
Links
Images
Abstract
A VLSI implementation system of a revised histogram equalization image based on statistical property comprises a block scanning element, a cumulative distribution function unit, a mean value, a variance analysis unit, a mapping table, an amendment element of grey mapping function, a grey mapping processing unit. The invention can improve the phenomenon of integral dark and integral bright caused by the uneven grayscale distribution of the digital image, which reasonably allocates distribution of the grayscale of the image and is suitable for the sequential image treatment processing of fingerprint identification.
Description
Technical Field
The invention relates to an image processing technology, in particular to a system and a method for realizing histogram gray level equalization based on statistical characteristic correction
Background
The histogram reflects the gray distribution information of an image, and therefore, the histogram is often used in image processing to analyze an image, and to perform gray adjustment and threshold generation. The purpose is as follows: threshold generation of binarization processing is realized, and an appropriate threshold is selected to binarize the image. Threshold generation in image segmentation is achieved. And extracting effective parts in the image and extracting the interested image content. Histograms are a classical way of use in terms of threshold generation due to the different purposes of use.
The existing histogram implementation system for fingerprint identification is mostly implemented by adopting a software mode:
one is an automatic Fingerprint verification system (AFIS) based on PC/Server, which uses the strong computing power of PC/Server CPU to implement traversal of whole Fingerprint image by using software, and then implements distribution analysis of gray value on the basis of the traversal, so as to implement histogram equalization processing. The method is mainly used for large-scale non-real-time use occasions. The method is suitable for the field with stronger specialty and has high use and maintenance cost. Due to the cost and user interface limitations, the usage approach is not practical for use in the consumer electronics field.
Still another is an AFIS based embedded system of embedded CPU (such as ARM, MIPS, etc.), which uses a software processing system based on real-time operating system to implement the processing of fingerprint images. The AFIS processing capability and speed of the embedded system are lower than those of an ASIC, and the embedded CPU, a real-time operating system and embedded software are higher in development cost, so that the fingerprint identification system cannot be used in a large quantity. But the use interface is friendly, and the method is suitable for consumer electronics, but the development cost cannot be popularized in a large amount, which is also a problem faced by the current fingerprint market.
One of the existing histogram processing methods for fingerprint identification is to process and analyze the entire fingerprint image by using a global histogram analysis method.
Due to the complex fingerprint acquisition environment in the fingerprint identification system and the poor consistency of signal conversion of various sensors to the finger, the local exposure imbalance of the fingerprint image is generated in the acquisition process, and the local information of the image can be ignored by using the histogram gray level processing mode of the global image, so that the processing of the fingerprint image is influenced.
The other is a single processing mode that directly employs classical histogram equalization. Histogram correction is not performed for the characteristics of the fingerprint image.
And when the image gray distribution acquired by the sensor does not aim at the characteristics of the fingerprint image, performing histogram correction. The histogram equalization is directly adopted, under the condition that the gray distribution is seriously uneven, the processed gray is combined, the gray level is reduced, the gray information disappears, the details of the fingerprint image are lost, and the feature point extraction of the subsequent operation is influenced.
Disclosure of Invention
The invention aims to solve the problems and discloses a histogram gray scale balancing VLSI implementation system and method based on statistical characteristic correction, which can improve the phenomenon that the whole image is dark or bright due to uneven gray scale distribution of a digital image, complete gray scale conversion processing, rationalize the gray scale distribution of the image and be more suitable for subsequent image processing of fingerprint identification.
A VLSI implementation system for histogram equalization image processing based on statistical property correction comprises the following units:
(1) The block scanning unit comprises a histogram generating unit and a statistical characteristic acquiring unit and is used for traversing image blocks by receiving image block data, finishing the check of gray distribution, constructing a histogram and traversing a primary image, finishing the detection of an image mean value, finishing the detection of an absolute mean difference value for the second time of traversal, and obtaining the mean value and the variance of the statistical characteristic of the image blocks;
(2) The cumulative distribution function unit is used for processing the sequence of the histogram according to the histogram of the image to obtain a cumulative distribution function describing the probability distribution of the gray value;
(3) The mean value and variance analysis unit is used for comparing the absolute mean difference of each pixel of the image block, finding out the maximum and minimum values of the absolute mean difference, taking the corresponding gray value as a threshold, carrying out threshold segmentation on the image, and respectively establishing gray mapping on the segmented image;
(4) The mapping table comprises a normalized cumulative distribution function unit and a gray mapping function unit, and is used for carrying out normalized processing on the cumulative distribution function, establishing a gray mapping sequence table with histogram equalization and establishing a gray mapping function;
(5) The correcting unit of the gray mapping function is used for correcting the gray mapping function of the gray balance of the square image by referring to the statistical gray mapping function and the minimum absolute mean difference and mean value of the image;
(6) And the gray mapping processing unit is used for inquiring the corresponding gray value by taking the pixel value of the input block image as an index to complete the gray conversion of the image.
The VLSI implementation system of histogram equalization image processing based on statistical property correction further comprises:
(21) The noise suppression and image blocking unit is used for dividing the image into a series of image small blocks and performing noise suppression processing on the image small blocks, so that the image quality is improved, and the difficulty of later-stage fingerprint feature point extraction is reduced;
(22) And the image cache reading controller is used for completing space access to the image, controlling the content of the image in real time and sequentially sending the content to the histogram generation unit at certain time intervals.
The histogram generating unit further includes a register having a gray level of 0 to 255 for storing the histogram.
The VLSI implementation system for histogram equalization image processing based on statistical characteristic correction further comprises a fingerprint image subsequent processing unit for the steps of fingerprint image binarization, refinement and feature point extraction.
The invention also discloses a VLSI implementation method of the histogram equalization image processing based on the statistical characteristic correction, which comprises the following steps:
step 3, analyzing and processing the input mean value and variance, and generating different gray level mapping sequence tables for different gray level ranges;
step 4, analyzing and calculating the cumulative distribution function and the normalized cumulative distribution function of the histogram, and establishing a gray mapping processing function;
and 5, converting the gray scale space according to the gray scale mapping processing table, thereby realizing uniform histogram balance.
And step 60, carrying out noise suppression processing on the small image blocks, improving the image quality and reducing the difficulty of later fingerprint feature point extraction.
The method for calculating the image variance in step 3 is as follows:
BlockAME[i][j]＝Block[i][j]-BlockMean
wherein, blockmean represents the image block mean, blockAME [ i ] [ j ] represents the image block variance,
w is the width of the image block, h is the height of the image block, i = 0-w-1 is a traversal variable in the horizontal direction, and j = 0-h-1 is a traversal variable in the vertical direction.
The histogram establishing method in the step 2 is as follows:
for(j＝0；j＜w；j++)
for(j＝0；j＜w；j++)
BlockHist[Block[i][j]]＝BlockHist[Block[i][j]]+1。
the gray mapping sequence table in step 3 is:
wherein: k is a gray variable, and the value range is as follows: 0 to i
i is the current gray level, and the value range is as follows: 0 to 255
cdf is cumulative distribution function
BlockHist [ k ], number of image pixels with a gray scale of k.
The gray mapping processing function in step 4 is:
NormCdf[i]＝cdf[i]×255÷(w×h)
wherein: cdf [ i ], the sum of the number of image pixels with gray values of 1 to i.
The invention has the following effects:
the histogram performs two functions, one is image enhancement based on modified histogram equalization, and the other is binarization threshold generation of the image.
And adopting a local histogram equalization method as an enhancement processing mode of the fingerprint image quality.
The histogram is corrected by analyzing the statistical properties (mean, variance) of the local processing blocks.
The realization method of the full hardware assembly line and the parallel processing in the chip realized by the VLSI technology.
A method for VLSI implementation in pipelined, parallel processing mode of histogram equalization modified based on statistical properties.
The chip working mode based on VLSI technology adopts a pipeline working mode for block processing, histogram detection and statistical detection are parallel processing modes, the running speed is high, the cost is low, and the method is suitable for popularization of consumer electronics.
Based on the detection of the statistical characteristics, the detail loss caused by the mergence of pixels occurring in the image processing of the histogram equalization is corrected.
The block-based image processing mode well considers noise interference caused by environmental factors of fingerprint image acquisition, and the block processing is favorable for optimizing a local enhancement processing mode.
Drawings
FIG. 1 is a block diagram of the system of the present invention;
FIG. 2 is a flow chart of the present invention;
FIG. 3 is a schematic representation of the original embodiment of the present invention;
fig. 4 is a histogram after the equalization process of the present invention.
Detailed Description
The invention discloses a VLSI implementation system for histogram equalization image processing based on statistical characteristic correction, which mainly comprises the following modules as shown in a system block Diagram (Diagram 1) of FIG. 1
1. The noise reduction and image blocking unit 1 is used for dividing an image into a series of image small blocks, and performing noise suppression processing on the image small blocks, so that the image quality is improved, and the difficulty of later-stage fingerprint feature point extraction is reduced.
2. Image Buffer read controller 2 (Image Buffer read controller) (not consistent with the noise reduction, image blocking unit, and Image Buffer read controller in the figure)
The controller is used for completing space access to the images, controlling the content of the images in real time and sending the content of the images out in sequence at certain time intervals. Data processing of the block is completed before the image data block is sent out each time, and then data statistics and histogram construction are realized by the following two functional modules (histogram generation unit and statistical characteristic mean and absolute mean difference detection.
3. The block scanning unit 3 includes a histogram generating unit 31 (Build histogram) and a statistical characteristic obtaining unit 32:
histogram generating unit 31 (Build histogram)
The image block data output by the functional module is received, the image block is traversed (scanning is completed on the whole image or the image block once), the gray distribution is checked, the image is stored in a register corresponding to the gray scale of 0-255, the construction of a histogram is completed after the image transmission is completed, and the construction technology is the same as that of the prior art.
Statistical property Mean and absolute Mean Detection 32 (Mean & AMS Detection)
The detection of the image mean value can be completed by traversing the image once, the detection of the absolute mean difference can be completed by traversing the image for the second time, the statistical characteristics of the image block are obtained, and preparation is made for the next statistical analysis.
4. Cumulative distribution function unit 4 (Cdf for histogram)
From the histogram of the image, the sequence of histograms is subjected to a cumulative distribution function (cdf for short) between 0 and 255, which completely describes the probability distribution of the grey values.
5. Mean and variance analysis unit 5 (Mean & AMS analyzer)
Comparing absolute mean errors (absolute mean error) of each pixel of the image block, finding out the maximum and minimum values of the absolute mean errors, taking the corresponding gray value as a threshold, performing threshold segmentation on the image, and respectively establishing gray mapping on the segmented image.
6. A mapping table 6 including a normalized cumulative distribution function unit and a grayscale mapping function unit
Normalized cumulative distribution function unit 61 (normaize Cdf for Histogram)
Normalizing the CDF to 0-255, establishing a histogram balanced gray mapping sequence table,
a Gray mapping function unit 62 (Rebuild Gray mapping function) is established.
7. Correction unit 7 of Gray mapping function (modification Gray mapping function)
And correcting the gray mapping table of gray balance of the square chart by referring to the statistical gray mapping table and the minimum absolute mean difference and mean value of the image. The gray value distribution of the equalized image is more reasonable and uniform, and the subsequent feature point extraction work is facilitated.
8. Gray Mapping process 8 (Gray Mapping)
Inputting a block image, and inquiring corresponding gray values by taking pixel values of the block image as indexes to finish gray scale conversion of the image.
9. Fingerprint image subsequent processing 9 (floating processing)
Binarization, refining, feature point extraction and the like.
The VLSI implementation method of the present invention is as follows, see FIG. 2, VLSI implementation workflow (Diagram 2)
And step 2, carrying out noise suppression processing on the small image blocks (S2), improving the image quality and reducing the difficulty of later fingerprint feature point extraction. The sequence of step 2 and step 1 can be interchanged.
And step 3, scanning the whole block Image through an Image Buffer read controller (S3), and sequentially outputting the Image according to pixel data (each pixel data width is one byte).
And 4, when the block image is read out, calculating the mean value of the block in real time, and calculating the variance of the image after the image traversal is finished. (S4)
BlockAME[i][j]＝Block[i][j]-BlockMean
Wherein, blockmean represents the image block mean,
BlockAME [ i ] [ j ] denotes the image block variance,
i =0 to w-1, is a drift variable in the horizontal direction,
j =0 h-1 is a traversal variable in the vertical direction,
w is the width of the image block,
h is the height of the image block.
for(j＝0；j＜w；j++)
for(j＝0；j＜w；j++)
BlockHi st[Block[i][j]]＝BlockHi st[Block[i][j]]+1；
And step 6, analyzing and processing the mean value and the variance input in the step 4 in the step, and generating different gray mapping functions for different gray ranges. (S6)
Wherein:
k is a gray variable, and the value range is as follows: 0 to i
i is the current gray level, and the value range is as follows: 0 to 255
cdf is cumulative distribution function
BlockHist [ k ], number of image pixels with a gray scale of k.
And 7, analyzing the histogram generated in the step 5, analyzing and calculating the cumulative distribution function and the normalized cumulative distribution function of the histogram, and establishing a gray scale mapping processing table. (S7).
NormCdf[i]＝cdf[i]×255÷(w×h)
Wherein:
w is the width of the image block,
h is the height of the image block,
cdf [ i ], the sum of the number of image pixels with gray values of 1 to i.
And 8, completing the histogram equalization operation according to the steps, and realizing the conversion of the gray scale space according to the gray scale mapping processing table (S9), thereby realizing the histogram equalization operation. (S8)
The above process is demonstrated below using a specific implementation example.
For example, a fingerprint image is divided into 32X32 blocks, the input fingerprint image is divided into 32X32 blocks, the image block with width and height of 32X32 is used as the unit block to be processed, the processing of the unit block is according to the pipeline processing mode, and the image unit block processing process is as follows:
1. the read control unit of the image processing reads out a 32 × 32 unit block from the buffer, and inputs the unit block to the statistical calculation steps (s 4, s 6) and the histogram detection steps (s 5, s 7) in parallel.
2. In the statistical property calculation step, the average value and the absolute average difference (variance) of the image unit block are calculated in s4, and then the average value and the absolute difference are analyzed in s 5. And output to s8 to correct the gray scale transform of the histogram.
3. In the histogram detection step, the histogram generation unit (s 5) is firstly entered, then the histogram analysis unit is entered, the histogram analysis is carried out, a preliminary gray mapping table is generated, and the detection data is output to s8.
4. And (5) after the data enter s8, adjusting the gray level histogram of the image block and correcting the gray level mapping table according to the detected mean value and absolute difference.
5. A gradation conversion operation is performed at s9, and a processed image is output.
Fig. 3 is an original histogram according to the present invention, and fig. 4 is a histogram after equalization processing according to the present invention. The abscissa represents the pixel gray value. The ordinate represents the number of pixels. The image processing method can improve the phenomenon that the whole image is dark or bright due to uneven gray distribution of the digital image through the equalization processing, complete the gray conversion processing, rationalize the gray distribution of the image and be more suitable for the subsequent image processing of fingerprint identification.
Claims (10)
1. A VLSI implementation system for histogram equalization image processing based on statistical property correction is characterized by comprising the following units:
(1) The block scanning unit comprises a histogram generation unit and a statistical characteristic acquisition unit and is used for traversing image blocks by receiving image block data, finishing the check of gray distribution, constructing a histogram and traversing a primary image, finishing the detection of an image mean value, finishing the detection of an absolute mean difference by traversing for the second time, and obtaining the statistical characteristic mean value and the variance of the image blocks;
(2) The cumulative distribution function unit is used for processing the sequence of the histogram according to the histogram of the image to obtain a cumulative distribution function describing the probability distribution of the gray value;
(3) The mean value and variance analysis unit is used for comparing the absolute mean difference of each pixel of the image block, finding out the maximum and minimum values of the absolute mean difference, taking the corresponding gray value as a threshold, carrying out threshold segmentation on the image, and respectively establishing gray mapping on the segmented image;
(4) The mapping table comprises a normalized cumulative distribution function unit and a gray mapping function unit, and is used for carrying out normalized processing on the cumulative distribution function, establishing a gray mapping sequence table with histogram equalization and establishing a gray mapping function;
(5) The correcting unit of the gray mapping function is used for correcting the gray mapping function of the gray balance of the square image by referring to the statistical gray mapping function and the minimum absolute mean difference and mean value of the image;
(6) And the gray mapping processing unit is used for inquiring the corresponding gray value by taking the pixel value of the input block image as an index to complete the gray conversion of the image.
2. The histogram equalization image processing VLSI implementation system based on statistical property modification of claim 1, comprising:
(21) The noise suppression and image blocking unit is used for dividing the image into a series of image small blocks and performing noise suppression processing on the image small blocks, so that the image quality is improved, and the difficulty of later-stage fingerprint feature point extraction is reduced;
(22) And the image cache reading controller is used for completing the space access to the image, controlling the content of the image in real time and sequentially sending the content to the histogram generation unit at certain time intervals.
3. A histogram equalization image processing VLSI implementation system based on statistical property modification as claimed in claim 2 wherein said histogram generation unit further comprises a register with a gray level of 0-255 for storing the histogram.
4. The VLSI implementation system for histogram equalization image processing based on statistical property modification of claim 1, further comprising a fingerprint image post-processing unit for fingerprint image binarization, thinning, feature point extraction steps.
5. A VLSI implementation method of histogram equalization image processing based on statistical characteristic correction is characterized by comprising the following steps:
step 1, performing blocking operation on an input fingerprint image, and dividing the whole image into a plurality of blocks as a minimum processing unit for image enhancement and extraction; traversing the whole block of image, and sequentially outputting the image according to pixel data;
step 2, carrying out mean calculation on the read block images in real time, and calculating the variance of the images after the images are traversed; establishing a histogram of an input image processing pixel block;
step 3, analyzing and processing the input mean value and variance, and generating different gray mapping sequence tables for different gray ranges;
step 4, analyzing and calculating the cumulative distribution function and the normalized cumulative distribution function of the histogram, and establishing a gray mapping processing function;
and 5, converting the gray space according to the gray mapping processing table, thereby realizing uniform histogram balance.
6. A method for VLSI implementation of statistical property correction-based histogram equalization image processing as claimed in claim 5, characterized in that step 1 or step 2 comprises the following steps
And step 60, carrying out noise suppression processing on the image small blocks, improving the image quality and reducing the difficulty of later fingerprint feature point extraction.
7. A method for VLSI implementation of statistical property correction-based histogram equalization image processing as claimed in claim 5, characterized in that the image variance in step 3 is calculated as follows:
BlockAME[i][j]＝Block[i][j]-BlockMean
wherein, blockMean represents the image block mean, blockAME [ i ] [ j ] represents the image block variance,
w is the width of the image block, h is the height of the image block, i = 0-w-1 is a traversal variable in the horizontal direction, and j = 0-h-1 is a traversal variable in the vertical direction.
8. A method for VLSI implementation of statistical property modified histogram equalization for image processing as claimed in claim 7 wherein the histogram creation in step 2 is as follows:
for(j＝0；j＜w；j++)
for(j＝0；j＜w；j++)
BlockHist[Block[i][j]]＝BlockHist[Block[i][j]]+1。
9. the method of claim 8 for VLSI implementation of histogram equalization image processing based on statistical property modification, wherein the gray map sequence table in step 3 is:
wherein: k is a gray variable, and the value range is as follows: 0 to i
i is the current gray level, and the value range is as follows: 0 to 255
cdf is cumulative distribution function
BlockHist [ k ], number of image pixels with a gray scale of k.
10. A method for VLSI implementation of statistical property correction-based histogram equalization image processing as claimed in claim 9 wherein the gray-level mapping processing function in step 4 is:
NormCdf[i]＝cdf[i]×255÷(w×h)
wherein: cdf [ i ], the sum of the number of image pixels with gray values of 1 to i.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2007101214807A CN101114340A (en) | 2007-09-06 | 2007-09-06 | VLSI realizing system and method of histogram equalization image processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2007101214807A CN101114340A (en) | 2007-09-06 | 2007-09-06 | VLSI realizing system and method of histogram equalization image processing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101114340A true CN101114340A (en) | 2008-01-30 |
Family
ID=39022673
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2007101214807A Pending CN101114340A (en) | 2007-09-06 | 2007-09-06 | VLSI realizing system and method of histogram equalization image processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101114340A (en) |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101510302A (en) * | 2009-03-25 | 2009-08-19 | 北京中星微电子有限公司 | Method and apparatus for enhancing image |
| CN101566994B (en) * | 2008-04-22 | 2011-02-16 | 王磊 | Image and video retrieval method |
| CN102023160B (en) * | 2009-09-15 | 2012-07-25 | 财团法人工业技术研究院 | Measure method of combustible quality based on image |
| CN102999907A (en) * | 2012-11-17 | 2013-03-27 | 中国水产科学研究院渔业机械仪器研究所 | Image palette threshold segmentation method used for chlorella automatic counting |
| CN103852018A (en) * | 2012-12-03 | 2014-06-11 | 西安元朔科技有限公司 | Electric transmission line icing thickness measuring algorithm based on image processing |
| CN104408432A (en) * | 2014-11-28 | 2015-03-11 | 中国电子科技集团公司第二十八研究所 | Infrared image target detection method based on histogram modification |
| CN106557729A (en) * | 2015-09-30 | 2017-04-05 | 日本电气株式会社 | For processing the apparatus and method of facial image |
| CN107077740A (en) * | 2014-11-07 | 2017-08-18 | 富川安可股份公司 | For the method and system for the speed for determining to move flow surface |
| CN108280836A (en) * | 2018-01-24 | 2018-07-13 | 中科创达软件股份有限公司 | A kind of image processing method and device |
| CN109840912A (en) * | 2019-01-02 | 2019-06-04 | 厦门美图之家科技有限公司 | The modification method of abnormal pixel and equipment is calculated in a kind of image |
| CN109858449A (en) * | 2019-02-01 | 2019-06-07 | 安谋科技（中国）有限公司 | Image processing method, device, medium and system |
| CN110570445A (en) * | 2019-09-09 | 2019-12-13 | 上海联影医疗科技有限公司 | Image segmentation method, device, terminal and readable medium |
| CN110751603A (en) * | 2019-09-26 | 2020-02-04 | 中国科学院长春光学精密机械与物理研究所 | Method and system for enhancing image contrast and terminal equipment |
| CN110895665A (en) * | 2018-09-12 | 2020-03-20 | 上海耕岩智能科技有限公司 | Coordinate transformation method for under-screen imaging, storage medium and electronic equipment |
| CN110969594A (en) * | 2019-11-29 | 2020-04-07 | 广东优世联合控股集团股份有限公司 | Image splicing method |
| CN112348763A (en) * | 2020-11-09 | 2021-02-09 | 西安宇视信息科技有限公司 | Image enhancement method, device, electronic equipment and medium |
| CN115908428A (en) * | 2023-03-03 | 2023-04-04 | 山东大学齐鲁医院 | Image processing method and system for adjusting finger retractor |
-
2007
- 2007-09-06 CN CNA2007101214807A patent/CN101114340A/en active Pending
Cited By (24)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101566994B (en) * | 2008-04-22 | 2011-02-16 | 王磊 | Image and video retrieval method |
| CN101510302A (en) * | 2009-03-25 | 2009-08-19 | 北京中星微电子有限公司 | Method and apparatus for enhancing image |
| CN101510302B (en) * | 2009-03-25 | 2014-05-21 | 北京中星微电子有限公司 | Method and apparatus for enhancing image |
| CN102023160B (en) * | 2009-09-15 | 2012-07-25 | 财团法人工业技术研究院 | Measure method of combustible quality based on image |
| CN102999907A (en) * | 2012-11-17 | 2013-03-27 | 中国水产科学研究院渔业机械仪器研究所 | Image palette threshold segmentation method used for chlorella automatic counting |
| CN103852018A (en) * | 2012-12-03 | 2014-06-11 | 西安元朔科技有限公司 | Electric transmission line icing thickness measuring algorithm based on image processing |
| CN107077740A (en) * | 2014-11-07 | 2017-08-18 | 富川安可股份公司 | For the method and system for the speed for determining to move flow surface |
| CN104408432A (en) * | 2014-11-28 | 2015-03-11 | 中国电子科技集团公司第二十八研究所 | Infrared image target detection method based on histogram modification |
| CN104408432B (en) * | 2014-11-28 | 2017-05-10 | 中国电子科技集团公司第二十八研究所 | Infrared image target detection method based on histogram modification |
| CN106557729A (en) * | 2015-09-30 | 2017-04-05 | 日本电气株式会社 | For processing the apparatus and method of facial image |
| CN106557729B (en) * | 2015-09-30 | 2021-12-21 | 日本电气株式会社 | Apparatus and method for processing face image |
| CN108280836A (en) * | 2018-01-24 | 2018-07-13 | 中科创达软件股份有限公司 | A kind of image processing method and device |
| CN108280836B (en) * | 2018-01-24 | 2020-08-28 | 中科创达软件股份有限公司 | Image processing method and device |
| CN110895665B (en) * | 2018-09-12 | 2023-03-31 | 上海耕岩智能科技有限公司 | Coordinate transformation method for under-screen imaging, storage medium and electronic equipment |
| CN110895665A (en) * | 2018-09-12 | 2020-03-20 | 上海耕岩智能科技有限公司 | Coordinate transformation method for under-screen imaging, storage medium and electronic equipment |
| CN109840912A (en) * | 2019-01-02 | 2019-06-04 | 厦门美图之家科技有限公司 | The modification method of abnormal pixel and equipment is calculated in a kind of image |
| CN109840912B (en) * | 2019-01-02 | 2021-05-04 | 厦门美图之家科技有限公司 | Method for correcting abnormal pixels in image and computing equipment |
| CN109858449A (en) * | 2019-02-01 | 2019-06-07 | 安谋科技（中国）有限公司 | Image processing method, device, medium and system |
| CN110570445B (en) * | 2019-09-09 | 2022-03-25 | 上海联影医疗科技股份有限公司 | Image segmentation method, device, terminal and readable medium |
| CN110570445A (en) * | 2019-09-09 | 2019-12-13 | 上海联影医疗科技有限公司 | Image segmentation method, device, terminal and readable medium |
| CN110751603A (en) * | 2019-09-26 | 2020-02-04 | 中国科学院长春光学精密机械与物理研究所 | Method and system for enhancing image contrast and terminal equipment |
| CN110969594A (en) * | 2019-11-29 | 2020-04-07 | 广东优世联合控股集团股份有限公司 | Image splicing method |
| CN112348763A (en) * | 2020-11-09 | 2021-02-09 | 西安宇视信息科技有限公司 | Image enhancement method, device, electronic equipment and medium |
| CN115908428A (en) * | 2023-03-03 | 2023-04-04 | 山东大学齐鲁医院 | Image processing method and system for adjusting finger retractor |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101114340A (en) | VLSI realizing system and method of histogram equalization image processing | |
| CN109118473B (en) | Angular point detection method based on neural network, storage medium and image processing system | |
| CN111681197A (en) | Remote sensing image unsupervised change detection method based on Siamese network structure | |
| CN111079764B (en) | Low-illumination license plate image recognition method and device based on deep learning | |
| CN110555382A (en) | Finger vein identification method based on deep learning and Wasserstein distance measurement | |
| CN109871845B (en) | Certificate image extraction method and terminal equipment | |
| CN107967471A (en) | A kind of table tool automatic identifying method based on machine vision | |
| CN113808180B (en) | Heterologous image registration method, system and device | |
| CN108805863B (en) | Method for detecting image change by combining deep convolutional neural network with morphology | |
| CN113313657A (en) | Unsupervised learning method and system for low-illumination image enhancement | |
| CN110969620A (en) | Method and device for detecting magnetic shoe ripple defects | |
| CN108334875A (en) | Vena characteristic extracting method based on adaptive multi-thresholding | |
| CN111899168B (en) | Remote sensing image super-resolution reconstruction method and system based on feature enhancement | |
| CN112884758B (en) | Defect insulator sample generation method and system based on style migration method | |
| CN110929562A (en) | Answer sheet identification method based on improved Hough transformation | |
| CN115908142A (en) | Contact net tiny part damage testing method based on visual recognition | |
| CN112330562A (en) | Heterogeneous remote sensing image transformation method and system | |
| CN112288726A (en) | Method for detecting foreign matters on belt surface of underground belt conveyor | |
| CN104200460A (en) | Image registration method based on images characteristics and mutual information | |
| CN102799880B (en) | For catching the apparatus and method of rolling fingerprint image | |
| CN105354823A (en) | Tree-ring image edge extraction and segmentation system | |
| CN114782822A (en) | Method and device for detecting state of power equipment, electronic equipment and storage medium | |
| CN112686851B (en) | Image detection method, device and storage medium | |
| CN107123105A (en) | Images match defect inspection method based on FAST algorithms | |
| CN111881922A (en) | Insulator image identification method and system based on significance characteristics |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20080130 |