CN105678795B - A kind of field shoe watermark image method of inspection - Google Patents
A kind of field shoe watermark image method of inspection Download PDFInfo
- Publication number
- CN105678795B CN105678795B CN201610119713.9A CN201610119713A CN105678795B CN 105678795 B CN105678795 B CN 105678795B CN 201610119713 A CN201610119713 A CN 201610119713A CN 105678795 B CN105678795 B CN 105678795B
- Authority
- CN
- China
- Prior art keywords
- point
- personal characteristics
- mark
- shoes
- image
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
Landscapes
- Image Analysis (AREA)
Abstract
The embodiment of the present invention provides a kind of field shoe watermark image method of inspection, including：Field shoe watermark image is pre-processed, determines shoes mark mark；Using the first personal characteristics point of shoes mark mark described in Corner Detection Algorithm multiple scale detecting；Detect the second individual character characteristic point of the connected region of the edge image of the shoes mark mark；First, second personal characteristics includes：Outer damage feature, repairing feature and the attachment feature of the shoes mark mark；The stabilization personal characteristics point of the shoes mark mark is determined according to the first personal characteristics point and the second individual character characteristic point；Shoes watermark image is distinguished according to the stable personal characteristics point.Solve the timeliness sex chromosome mosaicism artificially examined and ambiguity problem.There is rotation, scaling, translation invariance for the Multi-scale corner detection strategy of personal characteristics, improve the accuracy of inspection result.
Description
Technical field
The present embodiments relate to image processing techniques and verification of traces technical field more particularly to a kind of field shoe impressions
As detection method.
Background technology
In the process of clear up a criminal case, the inspection and analysis of live shoes watermark image play detection work particularly significant
Effect.Automatic inspection for live trace, forms full ripe solution not yet both at home and abroad.
At this stage, in China's verification of traces work, by the effort of many experts, a set of exclusive technology body has been formd
System achieves many scientific and technological achievements advanced in the world, has cracked many criminal cases in practice at many aspects of trace research
Part.However verification of traces but because itself theory, using etc. a variety of factors limitation, cause its investigation put into practice in application above deposit
In numerous difficulties.It is mainly manifested in traditional manual manipulation method and seriously constrains verification of traces technology in scouting is solved a case
Effect.It excessively relies on expert and artificial calibration feature takes and there are ambiguity.
Invention content
The embodiment of the present invention provides a kind of field shoe watermark image method of inspection, to overcome above-mentioned technical problem.
The field shoe watermark image method of inspection of the present invention, including：
Field shoe watermark image is carried out to pre-process determining shoes mark mark；
Using the first personal characteristics point of shoes mark mark described in Corner Detection Algorithm multiple scale detecting；
Detect the second individual character characteristic point of the connected region of the edge image of the shoes mark mark；Described first, second
Property feature includes：Outer damage feature, repairing feature and the attachment feature of the shoes mark mark；
The stabilization individual character of the shoes mark mark is determined according to the first personal characteristics point and the second individual character characteristic point
Characteristic point；
Shoes watermark image is distinguished according to the stable personal characteristics point.
Further, described that field shoe watermark image is pre-processed, including：
Remove the ambient noise of field shoe watermark image；
The bianry image of shoes mark mark is extracted from the field shoe mark image using trace partitioning algorithm.
Further, the first personal characteristics point using shoes mark mark described in Corner Detection Algorithm multiple scale detecting,
Including：
The diameter for setting the first personal characteristics point, window is determined according to the diameter；
The corresponding first personal characteristics point candidate point of the bianry image of shoes mark mark is determined according to the window, described first
Personal characteristics point candidate point is whole personal characteristics points including shoes mark mark；
The personal characteristics point except shoes mark mark in the first personal characteristics point candidate point is removed, the first individual character spy is obtained
Sign point.
Further, described that whole of the bianry image of shoes mark mark including shoes mark mark is determined according to the window
Personal characteristics point, including：
Judge whether the angle point number in the window meets given threshold, if so, determining that the angle point is described
Property characteristic point, if not, it is determined that the angle point is not personal characteristics point.
Further, the fringe region of the connection shoes mark mark detects the second individual character feature of the shoes mark mark
Point, including：
It is connected to the two-value shoes mark mark fringe region and obtains complete area, and mark the complete area；
The perimeter of the complete area is calculated, if the perimeter meets given threshold, it is determined that the complete area center
Point is second individual character characteristic point.
Further, described that the shoes mark is determined according to the first personal characteristics point and the second individual character characteristic point
The stabilization personal characteristics point of mark, including：
Merge the first personal characteristics point and the second individual character characteristic point obtains stablizing personal characteristics point；
Stablize personal characteristics point candidate point according to the screening to obtain stablizing personal characteristics point.
Further, described special according to the non-similarity screening stable individual character between the stable personal characteristics point candidate point
Sign point candidate point, including：
Similar stable personal characteristics point candidate point in the stable personal characteristics point candidate point is deleted, obtains stablizing individual character
Characteristic point.
Further, described that shoes watermark image is distinguished according to the stable personal characteristics point, including：
Centered on the stable personal characteristics point, a diameter of length of side of the stable personal characteristics point determines square
Image block；
Shoes watermark image is distinguished according to the square image blocks.
The present invention first by analyze characteristic image to be extracted the characteristics of, to image preprocessing；Then it is main to imitate human eye
The detection process sensitive to the particular geometry of personal characteristics in image, for the apparent particular geometry of angle point, more
Nahoscale-level detects angle point, for the unconspicuous particular geometry of angle point, meets size requirements in the detection of edge image level
Particular geometry；The personal characteristics point extracted is finally integrated, screening conditions are arranged according to verification of traces priori, it is right
The personal characteristics point detected is further screened, and final personal characteristics is obtained.
The present invention is directed to live mark image, imitates human eye detection process, and automatic detection reflection shoes are distinguished during wearing
In the outer sole damage characteristic of other shoes, feature and attachment feature are repaired, feature point extraction is divided into two levels, respectively
Extraction.Automatic detection is had to geometric distortions such as rotation, scaling, translations not by the external feature of the introducings such as the behavioural habits of people
Sensitive characteristic.Automatic Detection and Extraction feature, need not manually mark picture feature, solve the timeliness sex chromosome mosaicism and two artificially examined
Adopted sex chromosome mosaicism.There is rotation, scaling, translation invariance for the Multi-scale corner detection strategy of personal characteristics, improve inspection
As a result accuracy.
Description of the drawings
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is this hair
Some bright embodiments for those of ordinary skill in the art without having to pay creative labor, can be with
Obtain other attached drawings according to these attached drawings.
Fig. 1 is the flow chart of the field shoe watermark image method of inspection of the present invention；
Fig. 2 is shoes watermark image the first personal characteristics point schematic diagram of the field shoe watermark image method of inspection of the present invention；
Fig. 3 is the shoes watermark image second individual character characteristic point schematic diagram of the field shoe watermark image method of inspection of the present invention；
Fig. 4 is that the shoes watermark image of the field shoe watermark image method of inspection of the present invention stablizes personal characteristics point schematic diagram；
Fig. 5 is that the shoes watermark image of the field shoe watermark image method of inspection of the present invention stablizes another schematic diagram of personal characteristics point；
Fig. 6 is the overall flow figure of the field shoe watermark image method of inspection of the present invention.
Specific implementation mode
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
The every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.
Fig. 1 is the flow chart of the field shoe watermark image method of inspection of the present invention, as shown in Figure 1, the method for the present embodiment can be with
Including：
Step 101 carries out field shoe watermark image to pre-process determining shoes mark mark；
Step 102, the first personal characteristics point using shoes mark mark described in Corner Detection Algorithm multiple scale detecting；
Step 103, the detection shoes mark mark edge image connected region second individual character characteristic point；Described
One, second individual character feature includes：Outer damage feature, repairing feature and the attachment feature of the shoes mark mark；
Step 104 determines the shoes mark mark according to the first personal characteristics point and the second individual character characteristic point
Stablize personal characteristics point；
Step 105 distinguishes shoes watermark image according to the stable personal characteristics point.
Further, described that field shoe watermark image is pre-processed, including：
Remove the ambient noise of field shoe watermark image；
The bianry image of shoes mark mark is extracted from the field shoe watermark image using trace partitioning algorithm.
Specifically, there is complicated background in usual field shoe watermark image, therefore be located in advance for field shoe watermark image
The characteristics of managing, analyzing characteristic image to be extracted, to image preprocessing, to remove the ambient noise of field shoe watermark image；This reality
It applies example pretreatment and shoes mark mark is extracted from the field shoe watermark image using image segmentation algorithm, output result is binary map
Picture, image segmentation algorithm is level-set segmentation method in the present embodiment.Then Harris Corner Detection Algorithm multiple scale detectings are used
Pretreated shoes mark mark obtains the first personal characteristics point, as shown in Fig. 2, imitating human eye mainly to personal characteristics in image
The detection process of particular geometry sensitivity detects angle point for the apparent particular geometry of angle point in multiple dimensioned level,
The second individual character characteristic point of the connected region of the edge image of secondary detection shoes mark mark, as shown in figure 3, finally according to the first individual character
Characteristic point and second individual character characteristic point determine the stabilization personal characteristics point of shoes mark mark, as shown in figure 4, to according to the stabilization
Property characteristic point distinguish shoes watermark image.Should have non-similarity, personal characteristics to have certain stability, individual character special between personal characteristics
On some scale effectively, personal characteristics has certain geometrical property to sign.The behavioural habits of analysis suspect body in trace
Existing feature analyzes the outer damage feature of trace, the embodiment of repairing feature and attachment feature on the image, setting screening item
Part removes pseudo- personal characteristics point.
The present embodiment field shoe watermark image method of inspection, the personal characteristics that this method detects can be to the identifications of suspect
Important reference is provided.First by analyzing characteristic image to be extracted the characteristics of, to image preprocessing；Then it is main to imitate human eye
The detection process sensitive to the particular geometry of personal characteristics in image, for the apparent particular geometry of angle point, more
Nahoscale-level detects angle point, and personal characteristics point is obtained after integrating screening angle point, for the unconspicuous particular geometry of angle point,
Edge image level detects the particular geometry for meeting size requirements, using geometry center as personal characteristics point；It is last whole
The personal characteristics point extracted is closed, screening conditions are arranged according to verification of traces priori, the personal characteristics detected is clicked through
Row further screening, obtains stable personal characteristics.The present embodiment field shoe prints mark image personal characteristics automatic detection algorithm,
Automatic detection is had insensitive to geometric distortions such as rotation, scaling, translations by the external feature of the introducings such as the behavioural habits of people
Characteristic.
Further, the first personal characteristics point using shoes mark mark described in Harris algorithm multiple scale detectings, packet
It includes：
The diameter for setting the first personal characteristics point, window is determined according to the diameter；
The corresponding first personal characteristics point candidate point of the bianry image of shoes mark mark is determined according to the window
The personal characteristics point except shoes mark mark in the first personal characteristics point candidate point is removed, the first individual character spy is obtained
Sign point.
Specifically, personal characteristics point is obtained after integrating screening angle point, for the unconspicuous particular geometry of angle point,
Edge image level, detection meets the particular geometry of size requirements, using geometry center as personal characteristics point；It is last whole
Close the personal characteristics point extracted.Multiple dimensioned level detect personal characteristics when, imitate human eye to things from the distant to the near, by obscuring
Become clearly cognitive process, Corner Detection is carried out respectively at multiple scales to image.Screening conditions are according to a large amount of numbers of observation
According to what is obtained, angle point number is more than two specially within the scope of particular geometry.Merge multiple dimensioned level and edge image layer
The personal characteristics point that face detects merges rule for the union for the personal characteristics point for taking two levels to detect.It is same for describing
The personal characteristics point of one personal characteristics merges.
Further, described that whole of the bianry image of shoes mark mark including shoes mark mark is determined according to the window
First personal characteristics point candidate point, including：
Judge whether the angle point number in the window is more than threshold value T1, if so, determining that the angle point is the individual character
Characteristic point, if not, it is determined that the angle point is not personal characteristics point.
Specifically, for field shoe mark mark the first personal characteristics of image zooming-out after pretreated, first individual character
Feature includes mainly the outer damage feature, repairing feature and attachment feature of the shoes mark mark, and the present embodiment is using more
Size measurement algorithm detects the first personal characteristics point of the shoes mark mark, and extraction process is as follows：
A, change Gaussian function variances sigma2(standard deviation sigma takes 5 values herein, corresponds to 5 scales, respectively value 1, and 2,3,4,
5) different gaussian kernel functions, are generated, with each layer gaussian kernel function of shoes mark mark image convolution, each layer convolution results are carried out
Harris Corner Detections (Corner Detection based on gray level image) obtain shoes mark mark image angle point.
B, it is screened with shoes print bianry image pair the first personal characteristics point handled in pretreatment, if detect first
Personal characteristics point is not stamped in shoes, removes the first personal characteristics point；
C, the diameter d of the first personal characteristics is set, the 1/60 of shoes mark mark effective length is usually chosen, according to the diameter
It determines window size (window is with the square of a diameter of length of side), the angle point that detected is screened, it is specific to sieve
Choosing method is：Judge whether the angle point number in the window is more than threshold value T1(T herein1Value 2), if so, determining the angle
Point is the first personal characteristics point, if not, it is determined that the angle point is not the first personal characteristics point.
Further, the fringe region of the connection shoes mark mark detects the second individual character feature of the shoes mark mark
Point, including：
It is connected to the two-value shoes mark mark fringe region and obtains complete area, and mark the complete area；
The perimeter of the complete area is calculated, if the perimeter meets given threshold, it is determined that the complete area center
Point is second individual character characteristic point.
Specifically, the fringe region for being connected to the shoes mark mark detects the second individual character characteristic point of the shoes mark mark,
Its extraction process is as follows：
A, two-value shoes mark mark edge detection obtains complete area, and marks the complete area；
B, the perimeter for calculating the complete area, if the perimeter meets given threshold T2, it is determined that the complete area
Central point is second individual character characteristic point.
Second individual character characteristic point predominantly detects the unconspicuous personal characteristics of corner feature, considers this category feature to be approximately round
Shape, characteristic diameter size illustrate in fisrt feature point detection process, shown in the acquiring method such as formula (1) of above-mentioned threshold value
π d=T2 (1)
Wherein, d is the diameter of the first personal characteristics, T2For the perimeter threshold of setting.
The feature sensitive to image border is examined according to human eye, personal characteristics point is detected in edge image level, to edge graph
As carrying out connected region detection, when connected region perimeter meets given threshold range, using connected region center as personal characteristics
Point.
Further, described that the shoes mark is determined according to the first personal characteristics point and the second individual character characteristic point
The stabilization personal characteristics point of mark, including：
Merge the first personal characteristics point and the second individual character characteristic point obtains stablizing personal characteristics point candidate point；
The stable personal characteristics point is screened according to the stable personal characteristics point candidate point.
Specifically, merge the first personal characteristics point and the second individual character characteristic point obtains stablizing personal characteristics point
Candidate point comprising following steps：
A, the first personal characteristics point and the second individual character characteristic point being merged, merging method is to take union,
It obtains stablizing personal characteristics point candidate point；
When B, stablizing personal characteristics point candidate's dot density more than threshold value in window described in multiple scale detecting level, it is believed that window
The interior same personal characteristics of feature point description, then merge personal characteristics point, and all personal characteristics points in window are merged into
A bit, the barycenter of personal characteristics point as in window；
Further, described special according to the non-similarity screening stable individual character between the stable personal characteristics point candidate point
Sign point candidate point comprising following steps：
It chooses small box gray matrix around stable personal characteristics point candidate point and describes the stable personal characteristics point candidate point,
The normalizated correlation coefficient between each stable personal characteristics point candidate point gray matrix in region is sought, if multiple stabilizations in region
Property characteristic point candidate point gray matrix between normalizated correlation coefficient value be more than given threshold T3(T is taken in implementation process3Equal to 0.5),
It then removes in region and has said similar stable personal characteristics point candidate point.
There should be non-similarity, for non-personal characteristics, to be gone if the different personal characteristics detected are similar between personal characteristics
Unless personal characteristics.
Further, described that shoes watermark image is distinguished according to the stable personal characteristics point, including：
Centered on the stable personal characteristics point, a diameter of length of side of the stable personal characteristics point determines square
Image block；
Shoes watermark image is distinguished according to the square image blocks.
Specifically, the personal characteristics finally extracted is centered on personal characteristics point, and personal characteristics diameter d is the length of side
Square image blocks.As shown in Figure 5.
The present invention first by analyze characteristic image to be extracted the characteristics of, to image preprocessing；Then it is main to imitate human eye
The detection process sensitive to the particular geometry of personal characteristics in image, for the apparent particular geometry of angle point, more
Nahoscale-level detects angle point, and for the unconspicuous particular geometry of angle point, the detection of edge image level meets size requirements
Particular geometry；The personal characteristics point extracted is finally integrated, screening conditions are arranged according to verification of traces priori, to inspection
The personal characteristics point measured is further screened, and final personal characteristics is obtained.The overall flow figure of the above method such as Fig. 6 institutes
Show.
The present invention prints mark image for field shoe, human eye detection process is imitated, during automatic detection reflection shoes are worn
It is different from the outer sole damage characteristic of other shoes, feature and attachment feature is repaired, feature point extraction is divided into two levels,
It extracts respectively.Automatic detection is had and is lost to geometry such as rotation, scaling, translations by the external feature of the introducings such as the behavioural habits of people
Very insensitive characteristic.Automatic Detection and Extraction feature, need not manually mark picture feature, solve the timeliness sex chromosome mosaicism artificially examined
And ambiguity problem.There is rotation, scaling, translation invariance for the Multi-scale corner detection strategy of personal characteristics, detecting
Aspect has apparent advantage.
During live verification of traces, by outer sole damage characteristic, repairing feature and the attachment feature of analyzing shoes
Reflection in the picture, features described above is reflected as being different from the geometric characteristic of shoe sole print feature in the picture, to reach
Detect the purpose of personal characteristics.Personal characteristics should embody following characteristics on the image：There should be non-similarity between personal characteristics, it is a
Property feature have certain stability, personal characteristics on some scale effectively, personal characteristics have certain geometrical property.And
During desk checking, human eye is sensitive to the particular geometry in image, and (particular geometry, which refers to, is different from flower in image
The geometry of line shape), and human eye is broadly divided into the description to shape and scale to the description of particular geometry.
Finally it should be noted that：The above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations；To the greatest extent
Present invention has been described in detail with reference to the aforementioned embodiments for pipe, it will be understood by those of ordinary skill in the art that：Its according to
So can with technical scheme described in the above embodiments is modified, either to which part or all technical features into
Row equivalent replacement；And these modifications or replacements, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution
The range of scheme.
Claims (4)
1. a kind of field shoe watermark image method of inspection, which is characterized in that including：
Field shoe watermark image is carried out to pre-process determining shoes mark mark, including：Remove the ambient noise of field shoe watermark image；Using
Trace partitioning algorithm extracts the bianry image of shoes mark mark from the field shoe watermark image；
Using the first personal characteristics point of shoes mark mark described in Corner Detection Algorithm multiple scale detecting；
The second individual character characteristic point of the connected region of the edge image of the shoes mark mark is detected, including：It is connected to the two-value shoes
Mark mark fringe region obtains complete area, and marks the complete area；The perimeter of the complete area is calculated, if the week
Length meets given threshold, it is determined that the complete area central point is second individual character characteristic point；First, second personal characteristics
Including：Outer damage feature, repairing feature and the attachment feature of the shoes mark mark；
The stabilization personal characteristics of the shoes mark mark is determined according to the first personal characteristics point and the second individual character characteristic point
Point, including：Merge the first personal characteristics point and the second individual character characteristic point obtains stablizing personal characteristics point candidate point；Root
The stable personal characteristics point candidate point is screened according to non-similarity between the stable personal characteristics point candidate point, including described in deletion
Similar stable personal characteristics point candidate point between stable personal characteristics point candidate point, obtains stablizing personal characteristics point；
Shoes watermark image is distinguished according to the stable personal characteristics point.
2. according to the method described in claim 1, it is characterized in that, described using shoes described in Corner Detection Algorithm multiple scale detecting
First personal characteristics point of mark mark, including：
The diameter for setting the first personal characteristics point, window is determined according to the diameter；
The corresponding first personal characteristics point candidate point of the bianry image of shoes mark mark is determined according to the window；
The personal characteristics point except shoes mark mark in the first personal characteristics point candidate point is removed, the first personal characteristics is obtained
Point.
3. according to the method described in claim 2, it is characterized in that, the binary map for determining shoes mark mark according to the window
As corresponding first personal characteristics point candidate point, including：
Judge whether the angle point number in the window meets given threshold, if so, determining that the angle point is that the individual character is special
Point is levied, if not, it is determined that the angle point is not personal characteristics point.
4. according to the method described in claim 1, it is characterized in that, described distinguish shoes impression according to the stable personal characteristics point
Picture, including：
Centered on the stable personal characteristics point, a diameter of length of side of the stable personal characteristics point determines square-shaped image
Block；
Shoes watermark image is distinguished according to the square image blocks.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610119713.9A CN105678795B (en) | 2016-03-02 | 2016-03-02 | A kind of field shoe watermark image method of inspection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610119713.9A CN105678795B (en) | 2016-03-02 | 2016-03-02 | A kind of field shoe watermark image method of inspection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105678795A CN105678795A (en) | 2016-06-15 |
| CN105678795B true CN105678795B (en) | 2018-09-21 |
Family
ID=56306439
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610119713.9A Active CN105678795B (en) | 2016-03-02 | 2016-03-02 | A kind of field shoe watermark image method of inspection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105678795B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110287843B (en) * | 2019-06-18 | 2022-01-25 | 清华大学深圳研究生院 | Cultural relic fingerprint area selection method |
| CN111000327A (en) * | 2019-10-24 | 2020-04-14 | 晋江市达亿经编织造有限公司 | Breathable mesh vamp capable of automatically eliminating shoe marks |
| CN111860500B (en) * | 2020-07-10 | 2024-03-19 | 大连海事大学 | Shoe stamp wearing area detection and edge tracing method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102867311A (en) * | 2011-07-07 | 2013-01-09 | 株式会社理光 | Target tracking method and target tracking device |
| CN104915932A (en) * | 2015-05-19 | 2015-09-16 | 中国电子科技集团公司第五十研究所 | Target feature-based holographic radar image preprocessing and target extraction method |
-
2016
- 2016-03-02 CN CN201610119713.9A patent/CN105678795B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102867311A (en) * | 2011-07-07 | 2013-01-09 | 株式会社理光 | Target tracking method and target tracking device |
| CN104915932A (en) * | 2015-05-19 | 2015-09-16 | 中国电子科技集团公司第五十研究所 | Target feature-based holographic radar image preprocessing and target extraction method |
Non-Patent Citations (4)
| Title |
|---|
| Local image Features for Shoeprint Image Retrieval;H. Su 等;《Proceedings of the British Machine Vision Conference 2007》;20070913;第1-10页 * |
| Partial shoeprint retrieval using multiple point-of-interest detectors and SIFT descriptors;S. Almaadeed 等;《Integrated Computer-Aided Engineering》;20150131;第22卷(第1期);第41页摘要，第44-45页第2节最后一段、第3节 * |
| 基于 Harris角点和SURF特征的遥感图像匹配算法;葛盼盼 等;《计算机应用研究》;20140731;第31卷(第7期);第2206页第3节 * |
| 斑点检测;Ronny;《http://www.cnblogs.com/ronny/p/3895883.html》;20140928;第3节 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105678795A (en) | 2016-06-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115375676B (en) | Stainless steel product quality detection method based on image recognition | |
| CN109377485B (en) | Machine vision detection method for instant noodle packaging defects | |
| Reid et al. | A semi-automated methodology for discontinuity trace detection in digital images of rock mass exposures | |
| CN103971098B (en) | Method for recognizing wall in house type image and method for automatically correcting length ratio of house type image | |
| CN101706445B (en) | Image processing mthod for beef marbling grain grade scoring | |
| CN110717888B (en) | Automatic identification method for intravascular Optical Coherence Tomography (OCT) vessel wall inner contour | |
| CN108364280A (en) | Structural cracks automation describes and width accurately measures method and apparatus | |
| CN103164692B (en) | A kind of special vehicle instrument automatic identification system based on computer vision and method | |
| Mo et al. | Deshadowing of urban airborne imagery based on object-oriented automatic shadow detection and regional matching compensation | |
| CN111915704A (en) | Apple hierarchical identification method based on deep learning | |
| CN104392240A (en) | Parasite egg identification method based on multi-feature fusion | |
| WO2019127619A1 (en) | Method and system of segmentation and identification of carpal bones, terminal, and readable storage medium | |
| CN106709484B (en) | Digital identification method of digital instrument | |
| CN105678795B (en) | A kind of field shoe watermark image method of inspection | |
| Dong et al. | A novel method for extracting information on pores from cast thin-section images | |
| CN112837290A (en) | Crack image automatic identification method based on seed filling algorithm | |
| CN109035227A (en) | The system that lung tumors detection and diagnosis is carried out to CT image | |
| Zhao et al. | The edge detection of river model based on self-adaptive Canny Algorithm and connected domain segmentation | |
| CN114549441A (en) | Sucker defect detection method based on image processing | |
| CN112734729A (en) | Water gauge water level line image detection method and device suitable for night light supplement condition and storage medium | |
| CN111597930A (en) | Coastline extraction method based on remote sensing cloud platform | |
| CN111145205B (en) | Pig body temperature detection method based on infrared image under multiple pig scenes | |
| CN107256421B (en) | Method for rapidly counting rice and wheat seeds | |
| CN108154116A (en) | A kind of image-recognizing method and system | |
| CN115311288B (en) | Method for detecting damage of automobile film |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |