CN105809087A - Radiation examination system and vehicle model template search method - Google Patents
Radiation examination system and vehicle model template search method Download PDFInfo
- Publication number
- CN105809087A CN105809087A CN201410838643.3A CN201410838643A CN105809087A CN 105809087 A CN105809087 A CN 105809087A CN 201410838643 A CN201410838643 A CN 201410838643A CN 105809087 A CN105809087 A CN 105809087A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- image
- feature
- transmission image
- template
- 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.)
- Granted
Links
Abstract
The invention discloses a radiation examination system and a vehicle model template search method thereof. The method comprises the steps of acquiring a transmittance image of an inspected vehicle; extracting features from the transmittance image; and searching a vehicle model template from a vehicle model template database via a feature matching method based on the extracted features. By adopting the method, an empty template of a vehicle model can be quickly searched and determined from the vehicle model database, the image judgment efficiency is improved by saving the procedure of manually selecting a template, auxiliary information is provided for image comparison software such as a subtraction algorithm and the like, and the method can also meet the requirements of applications for vehicle model recognition and the like.
Description
Technical field
Embodiments of the invention relate to the image recognition in radiation image and retrieval technique, in particular in radiation image-forming system, image information according to vehicle to be measured, fast search the method for determining the vehicle template image matched with its model in Standard of vehicle data base.
Background technology
The retrieval of vehicle template refers to that automatic searching database obtains the technology of the empty wagons template image of this vehicle models, belongs to the content category of image retrieval according to the vehicle image information collected.Image retrieval problem is born from 20 century 70s, develop at present, relatively the way of main flow is divided into two kinds: one retrieval being based on text, namely from aspect index images such as image name, picture size, compression type, author, ages, then with key word form query image storehouse, but it is limited to the limitation of vocabulary itself, easy ambiguity, update difficulty, constantly update the environment of change unsuitable for data base；Another kind is based on the retrieval of picture material, namely to image vision content, is analyzed such as color of image, textural characteristics, profile information etc. and retrieves.Its advantage is that the objective visual information comprised according to image itself is sub as the description of retrieval, it is not necessary to human intervention and explanation, has good real-time and extensibility.
Along with developing rapidly of multimedia technology and Internet network, the appearance of mass-memory unit and digitizer and extensive use, image, video data present the growth trend of geometric techniques, then occur in that jumbo image/video data storehouse.General text search engine can not meet user's request, and therefore, how retrieving useful image quickly and efficiently from large-scale image library has become the focus of research both at home and abroad.And as solving the key technology of this problem, CBIR technology has become the focus of domestic and international extensive concern, but these application are only limited to the visible images use at internet arena mostly, and in safe examination system, by use radioscopy image, image zooming-out information is carried out template retrieve this specific area, there is presently no open source literature targetedly.
Summary of the invention
In view of one or more problems of the prior art, it is proposed that a kind of radiation checking system and vehicle template search method, it is possible to assist security staff to search vehicle.
In one aspect of the invention, it is proposed that the vehicle template search method in a kind of radiation checking system, including step: obtain the transmission image of examined vehicle；Feature extraction is carried out for described transmission image；From vehicle template database, vehicle template is retrieved based on the feature extracted by the method for characteristic matching.
According to some embodiments, the step carrying out feature extraction for described transmission image includes: the transmission image of described vehicle is divided into headstock, vehicle body and three sub-blocks of the tailstock；Respective feature is extracted for described three sub-blocks；Combine described three respective features of sub-block, as the feature of described transmission image.
According to some embodiments, above-mentioned method further comprises the steps of: the striped removed in described transmission image, obtains the transmission image that change of background is mild；The transmission image eliminating striped is sought gradient, and quantifies gradient map, remove the impact of little Fluctuation of gradient；Quantify, in gradient map, to ask the maximum continuum of level, upright projection, i.e. maximum vehicle location therein in two-value；Take the minimum rectangle position at vehicle place in described transmission image as pending subimage, remove ambient air part；The size of vehicle subimage is zoomed in and out.
According to some embodiments, detect characteristic point for described transmission image by intensive sampling mode, extract feature descriptor；All characteristic descriptor sets are combined together, utilize feature clustering algorithm to generate dictionary；Characteristic information in statistical analysis dictionary, obtains corresponding to the numerical value vector of each image feature in data base.
According to some embodiments, the personnel that are operated by are local manually to be added vehicle template or updates model data storehouse by network.
According to some embodiments, using trie tree structure construction feature retrieval model, according to the equipment source of image difference, the tagsort that will extract, corresponding different device number sets up multiple trie tree.
According to some embodiments, quick retrieval is used to search out N number of vehicle template image most like with it in the trie tree corresponding with the device number of this image.
In another aspect of this invention, it is proposed that a kind of radiation image-forming system, including scanning subsystem, it is thus achieved that the transmission image of examined vehicle；Retrieval subsystem, carries out feature extraction for described transmission image, and retrieves vehicle template by the method for characteristic matching from vehicle template database based on the feature extracted.
In scheme as above, carry out feature extraction by treating the radiation image of measuring car, and retrieve in the Standard of vehicle view data in model data storehouse, thus obtaining the same model vehicle template image mated with vehicle model to be measured.The auxiliary information that this result is checked as dolly, by saving the operation of artificial selection's template.Additionally, the program is further vehicle cab recognition, subtraction algorithm, image provides strong support than equity, what substantially increase dolly examination flow process sentences figure efficiency.
Accompanying drawing explanation
In order to be more fully understood that the present invention, will describe the present invention according to the following drawings:
Fig. 1 is the structural representation of radiation checking system according to an embodiment of the invention；
Fig. 2 is the flow chart describing car modal search method according to an embodiment of the invention；
Fig. 3 is the example describing the process that image carries out feature extraction in the method shown in Fig. 2.
Fig. 4 describes the schematic diagram according to the example retrieving vehicle template in embodiments of the invention；
Fig. 5 is the schematic diagram describing the process that the transmission image to vehicle carries out pretreatment；And
Fig. 6 is according to the embodiments of the invention schematic diagram to the example that vehicle template is updated.
Detailed description of the invention
Specific embodiments of the invention are described more fully below, it should be noted that the embodiments described herein is served only for illustrating, is not limited to the present invention.In the following description, in order to provide thorough understanding of the present invention, elaborate a large amount of specific detail.It will be apparent, however, to one skilled in the art that: these specific detail need not be adopted to carry out the present invention.In other instances, in order to avoid obscuring the present invention, do not specifically describe known structure, material or method.
In entire disclosure, " embodiment ", " embodiment ", " example " or mentioning of " example " are meaned: the special characteristic, structure or the characteristic that describe in conjunction with this embodiment or example are involved at least one embodiment of the invention.Therefore, it is not necessarily all referring to same embodiment or example in the phrase " in one embodiment " of each local appearance of entire disclosure, " in an embodiment ", " example " or " example ".Furthermore, it is possible to any suitable combination and/or sub-portfolio by specific feature, structure or property combination in one or more embodiments or example.Additionally, it should be understood by one skilled in the art that term "and/or" used herein includes any and all combination of one or more relevant project listed.
For the problems of the prior art, the present invention proposes a kind of method that security staff of auxiliary carries out image template retrieval automatically.According to the method, in vehicle safety check process, it is possible to according to the feature proposed in the transmission image of checked vehicle, automatically recall from Standard of vehicle data base and the X-ray scanning Prototype drawing of the same model of vehicle to be checked, compare for security staff.So, by the ray scanning data of vehicle to be measured, characteristic information extraction, searching database, automatically chooses corresponding vehicle template image, eliminates the process that security staff manually searches one by one, there is provided for follow-up examination further and support, strengthen automaticity and the real-time of examination flow process.
Fig. 1 is the structural representation of radiation checking system according to an embodiment of the invention.As it is shown in figure 1, inspection system according to embodiments of the present invention relates to the technology automatically choosing vehicle template image in the safety inspection technology of X-radiation imaging, particularly compact car entrainment Automatic Measurement Technique.
System as shown in Figure 1 includes ID acquiring unit 110, radiation image-forming system 150, storage device 120, graphics processing unit 140 and display device 130.
In certain embodiments, ID acquiring unit 110 obtains the unique identifying number of examined vehicle.Such as this ID acquiring unit 110 includes video camera, is used for catching the license plate image of described examined vehicle；And recognition unit, it is used for the license plate number from vehicle examined described in license plate image identification.In other embodiments, ID acquiring unit 110 includes reader, reads the ID of described examined vehicle from the radio-frequency (RF) tag entrained by described examined vehicle.
Examined vehicle is carried out X-ray scanning by radiation image-forming system 150, obtains the radioscopic image of examined vehicle.Storage device 120 stores described radioscopic image and vehicle template database.
Graphics processing unit 140 retrieves the vehicle template corresponding with this vehicle from vehicle template database, it is determined that the variable domain between the radial transmission image and the template image that obtain.Display device 130 presents described variable domain to user.
Such as, when checking in when there being dilly to need, corresponding dilly is identified by ID acquiring unit 110 by vehicle recognition unit, generates the unique ID of software system and this dilly, such as license plate number.This vehicle unique ID is unique mark that this dilly is reached a standard in this software system.This mark ID can the data that generate for this dilly of software system, it is also possible to by identifying the license plate number of this vehicle, current software system is identified by license plate number.
When software system obtains after this dilly uniquely identifies (such as license plate number) by vehicle recognition unit, this dilly can be carried out x-ray scanning by radiation image-forming system 150, generates the x-ray scanning image of this dilly after respective algorithms processes.After x-ray scanning image generates, meeting and this dilly unique ID are associated, and are sent to display device 150 and display.
Graphics processing unit 140 is responsible for retrieving for template base, retrieves the template image corresponding with dilly to be checked.Determine the variable domain between the radial transmission image and template image obtained.Display device 130 presents described variable domain to user.
Such as, graphics processing unit (retrieval subsystem) carries out feature extraction for the transmission image of vehicle, the method then passing through characteristic matching retrieves vehicle template based on the feature extracted from vehicle template database, it is then determined that the variable domain between the transmission image obtained and vehicle template.
Fig. 2 is the flow chart describing car modal search method according to an embodiment of the invention.As in figure 2 it is shown, in step S21, obtained the transmission image of examined vehicle by radiation image-forming system 150.Then, in step S22, carry out you to feature extraction for transmission image.In step S23, from vehicle template database, detect vehicle template by the method for characteristic matching based on the feature extracted.Then, for instance determined the variable domain obtained between transmission image and vehicle template by amplifying methods such as outline.
Model data storehouse stores the template image of various vehicle.As shown in Figure 4, the standard picture of various vehicles carried out pretreatment, feature extraction and then set up model.In vehicle retrieving to be measured, first carry out vehicle image feature extraction to be measured, then carry out Model Matching.It addition, in certain embodiments, manually added by safety check operator or cloud platform central database network updates, it is possible to input new vehicle image and update retrieval model.
Such as, grid of reference data " 2014 up-to-date Specifications ", all vehicles in domestic upper road (including large-scale, pick up etc.) totally 178 brands, 1474 kinds of vehicles.If calculated by difference configuration, then there are about 2.5 ten thousand kinds.In the present invention, model data storehouse is mainly calculated by vehicle, gets rid of uncommon vehicle, and the scale of beginning is at about 500.In use, the data of new model, rare vehicle are supplemented at any time.
In addition, owing to the different radiogenic energy of scanning equipment/dosage is different, detector size is different, the radiation image that same vehicle scans under different equipment is not quite similar, when carrying out vehicle coupling, it is possible to vehicle to be measured is retrieved with in the identical vehicle template of the source device of this image with data base.Therefore except obtaining radiation image data, the information such as source device number and the energy/dosage that can also preserve this image, and with this, image in model data storehouse is classified in later retrieval flow process, testing image by preferential with and the identical Standard of vehicle data of its own facility information carry out retrieving and mating.
Due to CBIR (ContentBasedImageRetrieval, CBIR) overcome text based retrieval need manual annotation and may result in ambiguous problem, there is the advantage that automaticity is high, extensibility is strong, therefore can adopt CBIR technology.
Fig. 5 is the schematic diagram describing the process that the transmission image to vehicle carries out pretreatment.Owing to car load scan image data is excessive, in order to ensure the real-time processed, it is possible to carry out scanogram cutting out car, take the minimum rectangle position at vehicle place in artwork as pending subimage, remove ambient air part；Then the size of these vehicle subimages being zoomed in and out, scaling coefficient can be adjusted according to the information such as characteristic point quantity of concrete image original size, extraction.
As it is shown in figure 5, first carry out scanogram cutting out car, using marginal information as Main Basis, it is judged that the position of vehicle in image.Processing mode divides three steps to be: first removes striped, obtains the image that change of background is mild；Image is asked gradient, and quantifies gradient map, remove little Fluctuation of gradient impact (such as removing the impact of the gradient lower than certain threshold value)；Quantify, in gradient map, to ask the maximum continuum of level, upright projection, i.e. maximum vehicle location therein in two-value.Cutting out car process schematic as it is shown in figure 5, wherein a is artwork, b is figure, the c after striped is gradient map, and d is for cutting out car result figure.After eliminating the air part of vehicle periphery, the minimum rectangle position at vehicle place in artwork is carried out follow-up process as subimage；If the subimage after sanction car is still relatively big (length is more than 500 pixels), then the size of this image being zoomed in and out, length scale is to (width is convergent-divergent at equal pace) about 500 pixels.
Fig. 3 is the example describing the process that image carries out feature extraction in the method shown in Fig. 2.To all vehicle image zooming-out features, feature extraction can be manually choose, it is also possible to is automatically or semi-automatically extract；Can be of overall importance, such as entire image, it is also possible to be for certain target, the subregion etc. in image.The method of feature extraction has a lot, those skilled in the art should be readily appreciated that, some alternate algorithm such as color histogram of feature extracting method, Scale invariant features transform matching algorithm (ScaleInvariantFeatureTransform can also be associated, SIFT), most small nut value similar district operator (SmallestUnivalueSegmentAssimilatingNucleus, SUSAN), Harris corner detection operator, Surf algorithm (SpeedUpRobustFeature) etc..Preferably, full figure is carried out block statistics by word bag (Bagofwords, the BOW) model using information retrieval field conventional in this patent.
Such as, by intensive sampling (dense) mode, the training image in model data storehouse being detected characteristic point, extract surf feature descriptor, these descriptors are the vectors of 64 dimensions one by one, and what they represented is the characteristic point of local invariant in image；All characteristic descriptor sets are combined together, utilize feature clustering algorithm to generate dictionary.The algorithm generating dictionary has multiple, and K as is well known ties up means clustering algorithm (K-Means) and K ties up core singular value decomposition algorithm (K-Singularvaluedecomposition, K-SVD) etc..The present embodiment is selected K-Means algorithm construction one comprise the dictionary of K classification；The number of times that in statistics dictionary, each word occurs in the picture, thus representing each image in data base and becoming a K dimension value vector.This numerical value vector is the BOW feature of this image.
Due to the unbalanced feature of deformation difference under scanning vehicle image diverse location, and the situation that vehicle diverse location concordance own is different (as trunk region differs greatly), the surf feature clustering of whole figure is had bigger error.According to experimentation, vehicle image is divided into 3*2 the identical sub-block of size, respectively headstock by Aspect Ratio by us, vehicle body and the tailstock, each block is extracted respectively its BOW feature, then combination of eigenvectors is got up, namely realize the BOW feature extraction mode of piecemeal.
Being high dimension vector owing to the feature of said extracted is generally mostly, be equivalent to the nearest neighbor search problem of higher dimensional space when carrying out characteristic matching, its amount of calculation is too big, has a strong impact on the speed of retrieval, it is possible to build a retrieval model to accelerate the process of characteristic matching.The accelerating algorithm of high dimensional feature coupling has a lot, those skilled in the art should be readily appreciated that, multidimensional trie tree (the k-dimensionaltree based on tree construction that some alternate algorithm are such as traditional can also be associated, kd-tree), based on the vector approximation method of multi-resolution entity, Hash (multiplerandomizedsub-vectorsquantizationhashing is quantified based on repeatedly random subvector, MRSVQH) Index Algorithm, maximum fractionation hyperplane priority algorithm (BestBinFirst, BBF) etc..Preferably, the present invention is directed to the feature of BOW model and adopt the known KD tree-model of structure, carry out characteristic matching by the fast nearest-neighbor search algorithm (Fastlibraryforapproximatenearestneighbors, FLANN) of open code.
It is, for example possible to use KD tree construction feature retrieval model, different according to the equipment source of image, by BOW tagsort obtained above, corresponding different device number sets up multiple trie tree.For another example, given vehicle image to be measured, first pass around Image semantic classification, then the BOW feature of testing image is extracted, using FLANN algorithm to search out N number of vehicle template image most like with it in the trie tree corresponding with the device number of this image, wherein the size of N can need change according to what user applied.
So, the process of vehicle to be measured retrieval includes vehicle image feature extraction to be measured, Model Matching.First pass through feature extracting method and extract testing image feature, then use above-mentioned high dimensional feature matching algorithm searching database, the retrieval model corresponding with the device number of this image searches out the vehicle template image most like with it.
In certain embodiments, the data of vehicle can increase year by year, constantly to data base's updating maintenance and renewal, the renewal of data base can be passed through both of which by the present invention: operator are local manually adds vehicle template, the renewal of model data storehouse cloud central site network, as shown in Figure 6.Such as, when model data storehouse updates, KD tree can be added according to the interpolation of data base or deletion action or delete node by algorithm, to complete the renewal to retrieval model.
Adopt the present invention method can in model data storehouse fast search determine the empty wagons template of this vehicle models, by saving the operation of artificial selection's template, improve and sentence figure efficiency, there is provided auxiliary information for image comparison software such as subtraction algorithm etc., the needs of the application such as vehicle cab recognition can also be served simultaneously.In the present invention, automatically retrieve the template image of vehicle to be measured, there is through actual verification good performance and practicality.
Above detailed description, by using schematic diagram, flow chart and/or example, has elaborated numerous embodiments of radiation checking system and method thereof.When comprising one or more function and/or operation in this schematic diagram, flow chart and/or example, it will be understood by those skilled in the art that each function in this schematic diagram, flow chart or example and/or operation can by various structures, hardware, software, firmware or substantially their combination in any individually and/or jointly to realize.In one embodiment, some parts of theme described in embodiments of the invention can be realized by special IC (ASIC), field programmable gate array (FPGA), digital signal processor (DSP) or other integrated forms.But, those skilled in the art will recognize that, some aspects of embodiments disclosed herein can realize in integrated circuits on the whole or partly equally, be embodied as on one or more computer run one or more computer programs (such as, be embodied as in one or more computer system run one or more programs), it is embodied as one or more programs of running on the one or more processors (such as, be embodied as on one or more microprocessors run one or more programs), it is embodied as firmware, or substantially it is embodied as the combination in any of aforesaid way, and those skilled in the art are according to the disclosure, will be provided with design circuit and/or the ability of write software and/or firmware code.In addition, it would be recognized by those skilled in the art that, the mechanism of theme described in the disclosure can be distributed as the program product of various ways, and regardless of the actual particular type being used for performing the signal bearing medium of distribution, the exemplary embodiment of theme described in the disclosure is all applicable.The example of signal bearing medium includes but not limited to: recordable-type media, such as floppy disk, hard disk drive, compact-disc (CD), digital universal disc (DVD), digital magnetic tape, computer storage etc.；And transmission type media, such as numeral and/or analogue communication medium (such as, optical fiber cable, waveguide, wired communications links, wireless communication link etc.).
Although exemplary embodiment describing the present invention with reference to several, it is to be understood that, term used is to illustrate and exemplary and nonrestrictive term.The spirit without deviating from invention or essence can be embodied as in a variety of forms due to the present invention, it is to be understood that, above-described embodiment is not limited to any aforesaid details, and should explain widely in the spirit and scope that appended claims limit, therefore fall into the whole changes in claim or its equivalent scope and remodeling all should be appended claims and contained.
Claims (8)
1. the vehicle template search method in radiation checking system, including step:
Obtain the transmission image of examined vehicle；
Feature extraction is carried out for described transmission image；
From vehicle template database, vehicle template is retrieved based on the feature extracted by the method for characteristic matching.
2. the method for claim 1, wherein carries out the step of feature extraction for described transmission image and includes:
The transmission image of described vehicle is divided into headstock, vehicle body and three sub-blocks of the tailstock；
Respective feature is extracted for described three sub-blocks；
Combine described three respective features of sub-block, as the feature of described transmission image.
3. the method for claim 1, further comprises the steps of:
Remove the striped in described transmission image, obtain the transmission image that change of background is mild；
The transmission image eliminating striped is sought gradient, and quantifies gradient map, remove the impact of little Fluctuation of gradient；
Quantify, in gradient map, to ask the maximum continuum of level, upright projection, i.e. maximum vehicle location therein in two-value；
Take the minimum rectangle position at vehicle place in described transmission image as pending subimage, remove ambient air part；
The size of vehicle subimage is zoomed in and out.
4. the method for claim 1, wherein detects characteristic point for described transmission image by intensive sampling mode, extracts feature descriptor；All characteristic descriptor sets are combined together, utilize feature clustering algorithm to generate dictionary；Characteristic information in statistical analysis dictionary, obtains corresponding to the numerical value vector of each image feature in data base.
5. the method for claim 1, the personnel that are wherein operated by are local manually to be added vehicle template or updates model data storehouse by network.
6. the method for claim 1, wherein uses trie tree structure construction feature retrieval model, different according to the equipment source of image, and the tagsort that will extract, corresponding different device number sets up multiple trie tree.
7. the method for claim 1, wherein uses quick retrieval to search out one or more vehicle template image most like with it from the trie tree corresponding with the device number of this image.
8. a radiation checking system, including:
Imaging subsystems, it is thus achieved that the transmission image of examined vehicle；
Retrieval subsystem, carries out feature extraction for described transmission image, and retrieves vehicle template by the method for characteristic matching from vehicle template database based on the feature extracted.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410838643.3A CN105809087B (en) | 2014-12-30 | 2014-12-30 | Radiation inspection system and vehicle model template retrieval method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410838643.3A CN105809087B (en) | 2014-12-30 | 2014-12-30 | Radiation inspection system and vehicle model template retrieval method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105809087A true CN105809087A (en) | 2016-07-27 |
| CN105809087B CN105809087B (en) | 2020-03-27 |
Family
ID=56980846
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410838643.3A Active CN105809087B (en) | 2014-12-30 | 2014-12-30 | Radiation inspection system and vehicle model template retrieval method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105809087B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108109455A (en) * | 2016-11-24 | 2018-06-01 | 同方威视技术股份有限公司 | Safety check analog training device, method and system |
| CN108959417A (en) * | 2018-06-08 | 2018-12-07 | 安徽智网信息科技有限公司 | A method of based on quick-searching template dynamic generation column |
| CN111612015A (en) * | 2020-05-26 | 2020-09-01 | 创新奇智(西安)科技有限公司 | Vehicle identification method and device and electronic equipment |
| US11117488B2 (en) * | 2018-06-06 | 2021-09-14 | Lyft, Inc. | Systems and methods for matching transportation requests to personal mobility vehicles |
| CN114170418A (en) * | 2021-11-30 | 2022-03-11 | 吉林大学 | Automobile wire harness connector multi-feature fusion image retrieval method by searching images through images |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1107136A1 (en) * | 1999-12-01 | 2001-06-13 | Hyundai Electronics Industries Co., Ltd. | Content-based image retrieval system and method for retrieving images using the same |
| CN101162204A (en) * | 2006-10-10 | 2008-04-16 | 同方威视技术股份有限公司 | Small vehicle holding articles automatic detection method based on radiation image-changing testing |
| KR20110048801A (en) * | 2009-11-03 | 2011-05-12 | 주식회사 한매 | X-ray image search automated storage system |
| CN104239299A (en) * | 2013-06-06 | 2014-12-24 | 富士通株式会社 | Three-dimensional model retrieval method and apparatus |
-
2014
- 2014-12-30 CN CN201410838643.3A patent/CN105809087B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1107136A1 (en) * | 1999-12-01 | 2001-06-13 | Hyundai Electronics Industries Co., Ltd. | Content-based image retrieval system and method for retrieving images using the same |
| CN101162204A (en) * | 2006-10-10 | 2008-04-16 | 同方威视技术股份有限公司 | Small vehicle holding articles automatic detection method based on radiation image-changing testing |
| KR20110048801A (en) * | 2009-11-03 | 2011-05-12 | 주식회사 한매 | X-ray image search automated storage system |
| CN104239299A (en) * | 2013-06-06 | 2014-12-24 | 富士通株式会社 | Three-dimensional model retrieval method and apparatus |
Non-Patent Citations (1)
| Title |
|---|
| 李尹岑等: "肝脏CT图像特征提取方法的研究及其在检索中的应用", 《影像技术》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108109455A (en) * | 2016-11-24 | 2018-06-01 | 同方威视技术股份有限公司 | Safety check analog training device, method and system |
| US11117488B2 (en) * | 2018-06-06 | 2021-09-14 | Lyft, Inc. | Systems and methods for matching transportation requests to personal mobility vehicles |
| CN108959417A (en) * | 2018-06-08 | 2018-12-07 | 安徽智网信息科技有限公司 | A method of based on quick-searching template dynamic generation column |
| CN108959417B (en) * | 2018-06-08 | 2022-03-29 | 安徽智网信息科技有限公司 | Method for dynamically generating column based on quick retrieval template |
| CN111612015A (en) * | 2020-05-26 | 2020-09-01 | 创新奇智(西安)科技有限公司 | Vehicle identification method and device and electronic equipment |
| CN111612015B (en) * | 2020-05-26 | 2023-10-31 | 创新奇智(西安)科技有限公司 | Vehicle identification method and device and electronic equipment |
| CN114170418A (en) * | 2021-11-30 | 2022-03-11 | 吉林大学 | Automobile wire harness connector multi-feature fusion image retrieval method by searching images through images |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105809087B (en) | 2020-03-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhao et al. | A surface defect detection method based on positive samples | |
| KR101917000B1 (en) | Methods and systems for inspecting goods | |
| Ma et al. | Automatic segmentation of the prostate on CT images using deep learning and multi-atlas fusion | |
| CN105809087A (en) | Radiation examination system and vehicle model template search method | |
| CN102693299B (en) | System and method for parallel video copy detection | |
| Gao et al. | Generative adversarial networks for road crack image segmentation | |
| CN105809088A (en) | Vehicle identification method and system | |
| CN106610969A (en) | Multimodal information-based video content auditing system and method | |
| WO2019238104A1 (en) | Computer apparatus and method for implementing classification detection of pulmonary nodule images | |
| CN106844524A (en) | A kind of medical image search method converted based on deep learning and Radon | |
| Al-asadi et al. | Object based image retrieval using enhanced SURF | |
| Duan et al. | Unsupervised pixel-level crack detection based on generative adversarial network | |
| CN110737788B (en) | Rapid three-dimensional model index establishing and retrieving method | |
| Putzu et al. | Invariant moments, textural and deep features for diagnostic MR and CT image retrieval | |
| Hu et al. | Object detection in hospital facilities: A comprehensive dataset and performance evaluation | |
| Fan et al. | Pulmonary nodule detection using improved faster R-CNN and 3D Resnet | |
| Samala et al. | Analysis of deep convolutional features for detection of lung nodules in computed tomography | |
| Shin et al. | A novel cascade classifier for automatic microcalcification detection | |
| Xu et al. | Initial classification algorithm for pavement distress images using features fusion | |
| Qin et al. | Classification of pavement crack types based on square bounding box diagonal matching method | |
| Wang et al. | Application of data mining technology in medical image processing | |
| Wu et al. | Similar image retrieval in large-scale trademark databases based on regional and boundary fusion feature | |
| Verma et al. | Retrieval of better results by using shape techniques for content based retrieval | |
| Abe et al. | Similarity retrieval of trademark images by vector graphics based on shape characteristics of components | |
| Kerami et al. | Classification of X-ray images using grid approach |
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 |