CN101639858A - Image search method based on target area matching - Google Patents
Image search method based on target area matching Download PDFInfo
- Publication number
- CN101639858A CN101639858A CN200910109384A CN200910109384A CN101639858A CN 101639858 A CN101639858 A CN 101639858A CN 200910109384 A CN200910109384 A CN 200910109384A CN 200910109384 A CN200910109384 A CN 200910109384A CN 101639858 A CN101639858 A CN 101639858A
- Authority
- CN
- China
- Prior art keywords
- target area
- image
- user
- similarity
- searched
- 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
The invention discloses an image search method based on target area matching, which comprises the following steps: a, selecting an image to be searched, and utilizing an edge detection algorithm to automatically partition the image to be searched; b, judging the partitioning effect, confirming a searched target area if a user satisfies the target area which is automatically partitioned, and selecting a target area that the user is interested in as a searched target from images that are not partitioned by self if the user does not satisfy the target area which is automatically partitioned; c, extracting characteristic values of the target area; d, taking the target area confirmed by the user as a template, using a template matching mode to carrying out template matching and similarity calculation in an image library, taking the searched image as a target when the similarity reaches a preset threshold value, or carrying out the template matching and the similarity calculation continuously until the image is covered completely; and e, sequentially displaying the searched result according to the similarity after the images in a database are searched according to requirements.
Description
Technical field
The present invention relates to image processing techniques, more particularly, relate to a kind of image search method based on target area matching.
Background technology
Fast development along with network technology, and the applying of multimedia technology, all there is every day thousands of picture to produce, how from the image library of magnanimity like this, finds the image of meeting consumers' demand quickly and accurately, become the key issue that field of image search presses for solution.
Traditional image retrieval mode adopts the text-based image retrieval technology.This technology is carried out text marking in the mode of artificial treatment to each width of cloth image before retrieving, describe picture material in the number of words scope of certain limitation, forms the mark database of describing image, to be labeled as key word, associates with image data base.During retrieval, in fact only the image labeling database is operated, carried out text matches, thereby reach the purpose that respective image is retrieved.Yet this text-based image retrieval technology exists serious problem.At first, present computer vision and artificial intelligence technology all can't mark image automatically, and must depend on manually image are made mark.This work is not only wasted time and energy, and manual mark is inaccurate often or incomplete, also has subjective deviation inevitably.That is to say that different people has different understanding methods to same width of cloth image, the difference of this subjective understanding will cause the mismatch error in the image retrieval.In addition, the abundant visual signature (color or texture etc.) that is comprised in the image often can't be described objectively with text.
In order to overcome these problems, CBIR technology (Content-Based Image Retrieval) is arisen at the historic moment.Promptly set up index and carry out similarity retrieval by the approximate measure between the characteristics of image vector according to color, texture, shape and their characteristic informations such as assemblage characteristic that image comprised.Yet, the most existing image retrieval algorithm all is to lay particular emphasis on to consider image overall information, fail effectively to utilize the local feature of image, though these methods have calculate simple, to translation with rotate advantages such as insensitive, but they can't describe picture material difference spatially, thus statistics similarity that can only the movement images overall situation, can not be on the object level similarity of movement images, thereby the semanteme of the image expression of feature similarity may be different, and recall precision is not high.Simultaneously, existing not only target selection mode of image search method underaction based on local message, and have the low problem of recall precision that characteristic index too much causes.
Therefore, how to solve the defective that is difficult to accurately express the user search intention in the conventional images retrieval, and improve recall precision and precision ratio is the technical barrier that needs to be resolved hurrily in the industry.
The invention provides a kind of image search method based on target area matching, effectively improved the precision ratio and the recall ratio of image retrieval, strengthened the dirigibility of image retrieval.
Summary of the invention
The purpose of this invention is to provide a kind of image search method based on target area matching, this method strengthens the precision ratio and the recall ratio of image retrieval when improving the image indexing system performance, and, easy to use.
The technical solution adopted for the present invention to solve the technical problems is: a kind of image search method based on target area matching is provided, and this method comprises the following steps:
A, image to be retrieved of selection utilize edge detection algorithm to treat retrieving images and cut apart automatically;
If b, judgement segmentation effect are the target area satisfaction of user to cutting apart automatically, the then target area of deterministic retrieval; If the user is dissatisfied to the target area of cutting apart automatically, then the user is not selecting interested target area as searching object in the split image voluntarily;
The eigenwert of c, extraction target area;
D, target area that the user is determined be as template, uses the template matches mode to carry out template matches in image library and similarity is calculated, when similarity reaches preset threshold value, then with this searched image as hit; Otherwise, proceed template matches and similarity and calculate, till image is covered fully;
E, the image in database are retrieved on request and are finished, and then show result for retrieval according to the similarity order.
Compared with prior art, the method for the present invention's proposition has the following advantages:
1, overcome the global statistics method and can not effectively improve the precision ratio and the recall ratio of image retrieval in the defective of the similarity of object level movement images;
2, not only the retrieval that can express the user on the mode more flexibly more exactly of choosing in the target area is intended to, and, choosing of characteristic index optimized, recall precision is improved greatly.
Description of drawings
The present invention will be described in detail below in conjunction with preferred embodiment and accompanying drawing, wherein:
Fig. 1 is the process flow diagram of the inventive method;
Fig. 2 is the schematic diagram of the inventive method coupling.
Embodiment
The image search method based on target area matching that the present invention proposes is cut apart automatically to image to be retrieved, or manually cut apart, the deterministic retrieval target area, extract color, texture and the shape facility value of image target area then, and these eigenwert indexs are optimized choose, adopt regional matching technique to carry out image retrieval at last.
The inventive method mainly is divided into the extraction of the choosing of target area, target area comprehensive characteristics and three processes of image retrieval of based target zone coupling, respectively the specific implementation of these processes is described below.
The image automated regional is cut apart and the specific implementation of choosing of area-of-interest: utilize the characteristics that always have the edge between two adjacent areas with different gray-scale values, the present invention uses the classical image partition method based on the border, with target in the image and target, or the border of target and background divides, and the user can select area-of-interest to retrieve as the target area.Because the edge is the discontinuous result of gray-scale value, this discontinuous derivative that utilizes detects easily.Simultaneously, consider the situation that the visually-perceptible of cut zone that dividing method obtains and human eye may there are differences, so whether the user can select interested target area in the split image voluntarily according to the satisfaction decision of automatic segmentation effect.
The specific implementation that the target area comprehensive characteristics extracts: after the user determines down the target area, extract the color, texture, shape information in zone eigenwert as the target area.
Color characteristic is with respect to further feature, and is highly stable, and for rotation, translation, dimensional variation, even various deformation is all insensitive, shows very strong robustness, and calculates simple.The present invention uses histogram to describe the most directly perceived this and tangible characteristics of image.Color histogram is a kind of method of color distribution in the presentation video, its transverse axis is represented color grade, the longitudinal axis is illustrated in has this color pixel shared ratio in entire image on some color grades, each scale in the histogram color space has been represented a kind of color in the color space.The present invention's selection meets HSV (Smith, the 1978) color model of people's visual response, adopts the higher cumulative histogram method of recall precision to come the color characteristic of token image.In the extraction of color characteristic, adopt unequal interval to quantize to colourity (Hue) in the HSV model, be that H is between [0-7], because the people is mainly determined by colourity H the sensation of color, in order to have reduced data volume and the complexity calculated, we only choose the color feature value of 8 characteristic quantities of colourity H as this width of cloth image, are designated as λ respectively
1, λ
2..., λ
8Though adopt the similarity between the histogram calculation image fairly simple, it can not reflect the space characteristics of object in the image.
Textural characteristics can be used to the spatial information in the image is carried out to a certain degree quantitative description.Image can be regarded the combination in different texture zone as, and texture is normally defined certain local property of image, is gray scale or the spatial coherence of color or the visual performance of gray scale and color space position variation of neighbor.Utilizing textural characteristics to carry out image retrieval is a kind of very effective means.It is a lot of to be used for the method for analysis image texture, but the method for analyzing texture roughly can be divided into following three kinds: statistic law, structure method and Spectrum Method.The present invention uses the gray level co-occurrence matrixes method in the statistic law to extract the textural characteristics of target area, and these characteristic variables comprise energy, moment of inertia, unfavourable balance square and entropy, are designated as λ respectively
9, λ
10, λ
11, λ
12Wherein, energy is the inhomogeneity tolerance of gradation of image; Moment of inertia is the amount of grey scale change in the dimensioned plan picture; The unfavourable balance square is local stationarity again, and it reflects the also local stability of image; Entropy is the tolerance of the quantity of information that has of image.
Shape facility is as the key character of object in the picture engraving and regional characteristics, it is the important means of describing high-rise visual information (as target, object etc.), and target, object are even more important to obtaining image, semantic, must have good shape facility to describe support with extraction algorithm.Shape facility comprises zone, major axes orientation, square, excentricity, circular rate, tangent angle etc.In addition, for figure, shape is its unique important feature.The edge of certain target is called profile in the image, and is an importance of Image Retrieval based on the retrieval of shape or profile, and the shape or the profile of the image that it can sketch the contours from the user are started with, and retrieve the similar image of shape in image library.The present invention uses the shape facility describing method based on rim detection to come shape facility is extracted, and square, excentricity, circular rate, tangent angle as the shape facility value, are designated as λ respectively
13, λ
14, λ
15, λ
16
The specific implementation of the image retrieval of based target zone coupling: the present invention uses template matching method to carry out image retrieval.Be about to the target area that the user determines and will be template, in searched image on translation and similarity coupling, when similarity reaches desirable threshold value, then this searched image is returned as hit.Otherwise, proceed translation and the work of similarity coupling, till image is covered fully.In retrieving, service range measure of the present invention carries out the coupling of similarity between the zone and calculates, and the distance definition of regional b and b ' is as follows:
Wherein, w
c, w
tAnd w
sThe weight of representative color feature, textural characteristics and shape facility, and w respectively
c+ w
t+ w
s=1.
As shown in Figure 1, the embodiment of image search method based on target area matching is as follows:
Step S100: searching system work, the user opens or selects a secondary image to be retrieved;
Step S101: utilize edge detection algorithm that image is cut apart automatically, with target in the image and target, or the border of target and background divides, and the user can select area-of-interest to retrieve as the target area.The algorithm of rim detection is a lot, adopts simple here and the reasonable Sobe1 operator of effect is finished rim detection;
Step S102: determine the target area that need are retrieved according to the satisfaction that the image automated regional is cut apart;
Step S103:, then choose the user's interest target area as searching object if the user is satisfied to the target area of cutting apart automatically;
Step S104: consider the situation that the visually-perceptible of cut zone that dividing method obtains and human eye may there are differences, if the user is dissatisfied to the target area of cutting apart automatically, then the user can not select interested target area as searching object in the split image voluntarily;
Step S105: extract the color, texture, shape information of target area eigenwert as the target area.Be specially: selection meets the hsv color model of people's visual response, adopts the higher cumulative histogram method of recall precision to come the color characteristic of token image.In the extraction of color characteristic, adopt unequal interval to quantize to colourity (Hue) in the HSV model, be that H is between [0-7], because the people is mainly determined by colourity H the sensation of color, in order to have reduced data volume and the complexity calculated, we only choose the color feature value of 8 characteristic quantities of colourity H as this width of cloth image, are designated as λ respectively
1, λ
2..., λ
8Simultaneously, use the gray level co-occurrence matrixes method in the statistic law to extract the textural characteristics of target area, these characteristic variables comprise energy, moment of inertia, unfavourable balance square and entropy, are designated as λ respectively
9, λ
10, λ
11, λ
12At last, use shape facility describing method to come shape facility is extracted, square, excentricity, circular rate, tangent angle as the shape facility value, are designated as λ respectively based on rim detection
13, λ
14, λ
15, λ
16
Step S106: use the template matches mode to retrieve, and the similarity of calculating and target area in image library, when similarity reaches preset threshold value, then with this searched image as hit;
Step S107: the image in database is retrieved on request and is finished, and then according to the similarity order, can be descending, or the mode of ascending order shows result for retrieval.
As shown in Figure 2, the template matches schematic diagram is: the target area that the user is determined is as matching template T, in searched image, with from left to right, order translation from top to bottom and similarity coupling, when similarity reached desirable threshold value, this returned this searched image as hit.Otherwise, proceed translation and the work of similarity coupling, till image is covered fully.With Fig. 2 is example, and this template is the rectangle of M * L pixel size, and template T overlays the search graph S upward translation of size for N * N pixel, and that block search figure under template covers is called subgraph S
Ij, (i j) is the coordinate of upper left corner picture point in S figure of this piece subgraph, and 0≤i≤N-M+1,0≤j≤N-L+1.16 comprehensive characteristics values and subgraph S according to template
IjCorresponding comprehensive characteristics value carry out similarity degree and calculate, establish b=T, b '=S
Ij, carry out similarity by following formula and calculate,
Wherein, w
c, w
tAnd w
sThe weight of representative color feature, textural characteristics and shape facility, and w respectively
c+ w
t+ w
s=1.
The present invention utilizes the image automated regional to cut apart the mode of choosing the target area that combines with area-of-interest, on the basis that the target area comprehensive characteristics is extracted, use the template matches mode to carry out picture search and similarity calculating, realized the image retrieval in based target zone, overcome the global statistics method and can not effectively improve the precision ratio and the recall ratio of image retrieval in the defective of the similarity of object level movement images.Simultaneously, comprehensive target area selection mode has strengthened the dirigibility and the convenience of image retrieval.
Claims (8)
1. an image search method based on target area matching is characterized in that, this method comprises the following steps:
A, image to be retrieved of selection utilize edge detection algorithm to treat retrieving images and cut apart automatically;
If b, judgement segmentation effect are the target area satisfaction of user to cutting apart automatically, the then target area of deterministic retrieval; If the user is dissatisfied to the target area of cutting apart automatically, then the user is not selecting interested target area as searching object in the split image voluntarily;
The eigenwert of c, extraction target area;
D, target area that the user is determined be as template, uses the template matches mode to carry out template matches in image library and similarity is calculated, when similarity reaches preset threshold value, then with this searched image as hit; Otherwise, proceed template matches and similarity and calculate, till image is covered fully;
E, the image in database are retrieved on request and are finished, and then show result for retrieval according to the similarity order.
2. the method for claim 1, it is characterized in that: edge detection algorithm utilizes gray-scale value to calculate among the step a.
3. the method for claim 1, it is characterized in that: the eigenwert of the target area of extracting among the step c comprises color and/or texture and/or shape information.
4. method as claimed in claim 3 is characterized in that: adopt color histogram to extract the color feature value of target area.
5. method as claimed in claim 3 is characterized in that: adopt statistic law, structure method or Spectrum Method to extract the textural characteristics value of target area.
6. method as claimed in claim 3 is characterized in that: adopt the gray level co-occurrence matrixes method in the statistic law to extract the textural characteristics of target area
7. method as claimed in claim 3 is characterized in that: adopt shape facility describing method to come shape facility is extracted based on rim detection, with square, excentricity, circular rate, tangent angle as the shape facility value.
8, the method for claim 1 is characterized in that: comprise that also the eigenwert index to extracting among the step c is optimized the step of choosing.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN200910109384A CN101639858A (en) | 2009-08-21 | 2009-08-21 | Image search method based on target area matching |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN200910109384A CN101639858A (en) | 2009-08-21 | 2009-08-21 | Image search method based on target area matching |
Publications (1)
Publication Number | Publication Date |
---|---|
CN101639858A true CN101639858A (en) | 2010-02-03 |
Family
ID=41614842
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN200910109384A Pending CN101639858A (en) | 2009-08-21 | 2009-08-21 | Image search method based on target area matching |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101639858A (en) |
Cited By (40)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102012939A (en) * | 2010-12-13 | 2011-04-13 | 中国人民解放军国防科学技术大学 | Method for automatically tagging animation scenes for matching through comprehensively utilizing overall color feature and local invariant features |
CN102236675A (en) * | 2010-04-30 | 2011-11-09 | 华为技术有限公司 | Method for processing matched pairs of characteristic points of images, image retrieval method and image retrieval equipment |
CN102592144A (en) * | 2012-01-06 | 2012-07-18 | 东南大学 | Multi-camera non-overlapping view field-based pedestrian matching method |
CN102831239A (en) * | 2012-09-04 | 2012-12-19 | 清华大学 | Method and system for building image database |
CN103019605A (en) * | 2012-12-26 | 2013-04-03 | 广东欧珀移动通信有限公司 | Gesture shutdown method of touch intelligent terminal, and intelligent terminal |
CN103294699A (en) * | 2012-02-24 | 2013-09-11 | 联想(北京)有限公司 | Method and electronic equipment for screening object |
CN103870569A (en) * | 2014-03-13 | 2014-06-18 | 北方信息控制集团有限公司 | Colorful animal image retrieval method based on content and colorful animal image retrieval system based on content |
CN104731952A (en) * | 2015-03-31 | 2015-06-24 | 努比亚技术有限公司 | Method and device for checking pictures |
CN104778458A (en) * | 2015-04-17 | 2015-07-15 | 浙江理工大学 | Textile pattern retrieval method based on textural features |
CN104915955A (en) * | 2015-05-27 | 2015-09-16 | 上海交通大学 | Image-segmentation-based picture searching method |
WO2016062159A1 (en) * | 2014-10-20 | 2016-04-28 | 网易(杭州)网络有限公司 | Image matching method and platform for testing of mobile phone applications |
CN103226582B (en) * | 2013-04-08 | 2016-08-03 | 哈尔滨工程大学 | A kind of medical image search method based on uncertain fixed point figure |
CN106096255A (en) * | 2016-06-03 | 2016-11-09 | 宁波大红鹰学院 | A kind of infrared diagnostics system and method |
CN106462943A (en) * | 2014-11-18 | 2017-02-22 | 谷歌公司 | Aligning panoramic imagery and aerial imagery |
CN106649487A (en) * | 2016-10-09 | 2017-05-10 | 苏州大学 | Image retrieval method based on interest target |
CN106897328A (en) * | 2015-12-21 | 2017-06-27 | 苏宁云商集团股份有限公司 | A kind of image search method and device |
CN106903075A (en) * | 2017-04-26 | 2017-06-30 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of vamp logo multi-directional visions detection method and system |
CN106980866A (en) * | 2016-10-24 | 2017-07-25 | 商汤集团有限公司 | A kind of footwear search method |
CN107392913A (en) * | 2017-08-01 | 2017-11-24 | 郑州云海信息技术有限公司 | A kind of image partition method and device |
CN108228761A (en) * | 2017-12-21 | 2018-06-29 | 深圳市商汤科技有限公司 | The customized image search method in support area and device, equipment, medium |
CN108959338A (en) * | 2018-03-30 | 2018-12-07 | 东易日盛家居装饰集团股份有限公司 | A kind of image search method and system |
CN109276819A (en) * | 2017-07-20 | 2019-01-29 | 株式会社东芝 | Information processing unit, information processing system and storage medium |
CN109348731A (en) * | 2016-10-14 | 2019-02-15 | 深圳配天智能技术研究院有限公司 | A kind of method and device of images match |
CN109791559A (en) * | 2016-06-28 | 2019-05-21 | 谷歌有限责任公司 | Promote use of the image as search inquiry |
CN109784210A (en) * | 2018-12-26 | 2019-05-21 | 佛山科学技术学院 | A kind of scheduling positioning identifying method, system, terminal device and medium |
CN110069648A (en) * | 2017-09-25 | 2019-07-30 | 杭州海康威视数字技术股份有限公司 | A kind of image search method and device |
CN110413824A (en) * | 2019-06-20 | 2019-11-05 | 平安科技(深圳)有限公司 | A kind of search method and device of similar pictures |
CN110688903A (en) * | 2019-08-30 | 2020-01-14 | 陕西九域通创轨道系统技术有限责任公司 | Obstacle extraction method based on camera data of train AEB system |
CN110852319A (en) * | 2019-11-08 | 2020-02-28 | 深圳市深视创新科技有限公司 | Rapid universal roi matching method |
CN110851636A (en) * | 2018-07-24 | 2020-02-28 | 杭州海康威视数字技术股份有限公司 | Image searching system, method and device |
CN110929057A (en) * | 2018-08-30 | 2020-03-27 | 深圳市蓝灯鱼智能科技有限公司 | Image processing method, device and system, storage medium and electronic device |
CN111325079A (en) * | 2018-12-17 | 2020-06-23 | 北京华航无线电测量研究所 | Road surface pit detection method applied to vehicle-mounted vision system |
WO2020124629A1 (en) * | 2018-12-19 | 2020-06-25 | 惠科股份有限公司 | Method for determining level of similarity between adjacent rows in display image and display device |
WO2020124631A1 (en) * | 2018-12-19 | 2020-06-25 | 惠科股份有限公司 | Method for determining similarity of adjacent rows in picture, and display apparatus |
WO2020133236A1 (en) * | 2018-12-28 | 2020-07-02 | 深圳迈瑞生物医疗电子股份有限公司 | Spinal imaging method and ultrasonic imaging system |
CN112256910A (en) * | 2020-10-23 | 2021-01-22 | 蓝普金睛(北京)科技有限公司 | Content-based shop identification image retrieval method |
CN112507921A (en) * | 2020-12-16 | 2021-03-16 | 平安银行股份有限公司 | Graph searching method and system based on target area, electronic device and storage medium |
CN112990366A (en) * | 2021-04-23 | 2021-06-18 | 视睿(杭州)信息科技有限公司 | Target labeling method and device |
CN113676775A (en) * | 2021-08-27 | 2021-11-19 | 苏州因塞德信息科技有限公司 | Method for implanting advertisement in video and game by using artificial intelligence |
CN113760415A (en) * | 2020-06-04 | 2021-12-07 | Oppo广东移动通信有限公司 | Dial plate generation method and device, electronic equipment and computer readable storage medium |
-
2009
- 2009-08-21 CN CN200910109384A patent/CN101639858A/en active Pending
Cited By (56)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102236675A (en) * | 2010-04-30 | 2011-11-09 | 华为技术有限公司 | Method for processing matched pairs of characteristic points of images, image retrieval method and image retrieval equipment |
CN102012939B (en) * | 2010-12-13 | 2012-11-14 | 中国人民解放军国防科学技术大学 | Method for automatically tagging animation scenes for matching through comprehensively utilizing overall color feature and local invariant features |
CN102012939A (en) * | 2010-12-13 | 2011-04-13 | 中国人民解放军国防科学技术大学 | Method for automatically tagging animation scenes for matching through comprehensively utilizing overall color feature and local invariant features |
CN102592144A (en) * | 2012-01-06 | 2012-07-18 | 东南大学 | Multi-camera non-overlapping view field-based pedestrian matching method |
CN103294699A (en) * | 2012-02-24 | 2013-09-11 | 联想(北京)有限公司 | Method and electronic equipment for screening object |
CN102831239B (en) * | 2012-09-04 | 2016-01-20 | 清华大学 | A kind of method and system building image data base |
CN102831239A (en) * | 2012-09-04 | 2012-12-19 | 清华大学 | Method and system for building image database |
CN103019605A (en) * | 2012-12-26 | 2013-04-03 | 广东欧珀移动通信有限公司 | Gesture shutdown method of touch intelligent terminal, and intelligent terminal |
CN103226582B (en) * | 2013-04-08 | 2016-08-03 | 哈尔滨工程大学 | A kind of medical image search method based on uncertain fixed point figure |
CN103870569A (en) * | 2014-03-13 | 2014-06-18 | 北方信息控制集团有限公司 | Colorful animal image retrieval method based on content and colorful animal image retrieval system based on content |
CN103870569B (en) * | 2014-03-13 | 2017-05-10 | 北方信息控制集团有限公司 | Colorful animal image retrieval method based on content and colorful animal image retrieval system based on content |
WO2016062159A1 (en) * | 2014-10-20 | 2016-04-28 | 网易(杭州)网络有限公司 | Image matching method and platform for testing of mobile phone applications |
CN106462943A (en) * | 2014-11-18 | 2017-02-22 | 谷歌公司 | Aligning panoramic imagery and aerial imagery |
CN104731952A (en) * | 2015-03-31 | 2015-06-24 | 努比亚技术有限公司 | Method and device for checking pictures |
CN104778458A (en) * | 2015-04-17 | 2015-07-15 | 浙江理工大学 | Textile pattern retrieval method based on textural features |
CN104778458B (en) * | 2015-04-17 | 2017-12-22 | 浙江理工大学 | A kind of textile pattern search method based on textural characteristics |
CN104915955A (en) * | 2015-05-27 | 2015-09-16 | 上海交通大学 | Image-segmentation-based picture searching method |
CN106897328A (en) * | 2015-12-21 | 2017-06-27 | 苏宁云商集团股份有限公司 | A kind of image search method and device |
CN106096255A (en) * | 2016-06-03 | 2016-11-09 | 宁波大红鹰学院 | A kind of infrared diagnostics system and method |
CN109791559B (en) * | 2016-06-28 | 2023-10-10 | 谷歌有限责任公司 | Promoting image processing apparatus as a means of use of search queries |
CN109791559A (en) * | 2016-06-28 | 2019-05-21 | 谷歌有限责任公司 | Promote use of the image as search inquiry |
CN106649487A (en) * | 2016-10-09 | 2017-05-10 | 苏州大学 | Image retrieval method based on interest target |
CN109348731B (en) * | 2016-10-14 | 2022-05-17 | 深圳配天智能技术研究院有限公司 | Image matching method and device |
CN109348731A (en) * | 2016-10-14 | 2019-02-15 | 深圳配天智能技术研究院有限公司 | A kind of method and device of images match |
CN106980866A (en) * | 2016-10-24 | 2017-07-25 | 商汤集团有限公司 | A kind of footwear search method |
CN106903075A (en) * | 2017-04-26 | 2017-06-30 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of vamp logo multi-directional visions detection method and system |
CN109276819A (en) * | 2017-07-20 | 2019-01-29 | 株式会社东芝 | Information processing unit, information processing system and storage medium |
CN109276819B (en) * | 2017-07-20 | 2021-04-06 | 株式会社东芝 | Information processing apparatus, information processing system, and storage medium |
US10769467B2 (en) | 2017-07-20 | 2020-09-08 | Kabushiki Kaisha Toshiba | Information processing apparatus, information processing system, and computer program product |
CN107392913A (en) * | 2017-08-01 | 2017-11-24 | 郑州云海信息技术有限公司 | A kind of image partition method and device |
CN110069648A (en) * | 2017-09-25 | 2019-07-30 | 杭州海康威视数字技术股份有限公司 | A kind of image search method and device |
CN108228761B (en) * | 2017-12-21 | 2021-03-23 | 深圳市商汤科技有限公司 | Image retrieval method and device supporting region customization, equipment and medium |
CN108228761A (en) * | 2017-12-21 | 2018-06-29 | 深圳市商汤科技有限公司 | The customized image search method in support area and device, equipment, medium |
CN108959338A (en) * | 2018-03-30 | 2018-12-07 | 东易日盛家居装饰集团股份有限公司 | A kind of image search method and system |
CN110851636A (en) * | 2018-07-24 | 2020-02-28 | 杭州海康威视数字技术股份有限公司 | Image searching system, method and device |
CN110851636B (en) * | 2018-07-24 | 2023-05-02 | 杭州海康威视数字技术股份有限公司 | Image searching system, method and device |
CN110929057A (en) * | 2018-08-30 | 2020-03-27 | 深圳市蓝灯鱼智能科技有限公司 | Image processing method, device and system, storage medium and electronic device |
CN111325079A (en) * | 2018-12-17 | 2020-06-23 | 北京华航无线电测量研究所 | Road surface pit detection method applied to vehicle-mounted vision system |
WO2020124631A1 (en) * | 2018-12-19 | 2020-06-25 | 惠科股份有限公司 | Method for determining similarity of adjacent rows in picture, and display apparatus |
US11381828B2 (en) | 2018-12-19 | 2022-07-05 | HKC Corporation Limited | Method and display device for determining similarity of adjacent rows of an image |
US11967130B2 (en) | 2018-12-19 | 2024-04-23 | HKC Corporation Limited | Method for determining similarity of adjacent rows in a picture and display device |
WO2020124629A1 (en) * | 2018-12-19 | 2020-06-25 | 惠科股份有限公司 | Method for determining level of similarity between adjacent rows in display image and display device |
CN109784210A (en) * | 2018-12-26 | 2019-05-21 | 佛山科学技术学院 | A kind of scheduling positioning identifying method, system, terminal device and medium |
WO2020133236A1 (en) * | 2018-12-28 | 2020-07-02 | 深圳迈瑞生物医疗电子股份有限公司 | Spinal imaging method and ultrasonic imaging system |
CN112654301A (en) * | 2018-12-28 | 2021-04-13 | 深圳迈瑞生物医疗电子股份有限公司 | Imaging method of spine and ultrasonic imaging system |
CN110413824A (en) * | 2019-06-20 | 2019-11-05 | 平安科技(深圳)有限公司 | A kind of search method and device of similar pictures |
CN110688903B (en) * | 2019-08-30 | 2023-09-26 | 湖南九域同创高分子新材料有限责任公司 | Barrier extraction method based on train AEB system camera data |
CN110688903A (en) * | 2019-08-30 | 2020-01-14 | 陕西九域通创轨道系统技术有限责任公司 | Obstacle extraction method based on camera data of train AEB system |
CN110852319A (en) * | 2019-11-08 | 2020-02-28 | 深圳市深视创新科技有限公司 | Rapid universal roi matching method |
CN113760415A (en) * | 2020-06-04 | 2021-12-07 | Oppo广东移动通信有限公司 | Dial plate generation method and device, electronic equipment and computer readable storage medium |
CN112256910A (en) * | 2020-10-23 | 2021-01-22 | 蓝普金睛(北京)科技有限公司 | Content-based shop identification image retrieval method |
CN112507921A (en) * | 2020-12-16 | 2021-03-16 | 平安银行股份有限公司 | Graph searching method and system based on target area, electronic device and storage medium |
CN112507921B (en) * | 2020-12-16 | 2024-03-19 | 平安银行股份有限公司 | Target area-based graphic searching method, system, electronic device and storage medium |
CN112990366B (en) * | 2021-04-23 | 2021-09-07 | 视睿(杭州)信息科技有限公司 | Target labeling method and device |
CN112990366A (en) * | 2021-04-23 | 2021-06-18 | 视睿(杭州)信息科技有限公司 | Target labeling method and device |
CN113676775A (en) * | 2021-08-27 | 2021-11-19 | 苏州因塞德信息科技有限公司 | Method for implanting advertisement in video and game by using artificial intelligence |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101639858A (en) | Image search method based on target area matching | |
CN102622420B (en) | Trademark image retrieval method based on color features and shape contexts | |
CN103186904B (en) | Picture contour extraction method and device | |
CN101789005A (en) | Image searching method based on region of interest (ROI) | |
CN102073748B (en) | Visual keyword based remote sensing image semantic searching method | |
CN102663391B (en) | Image multifeature extraction and fusion method and system | |
Jasiewicz et al. | Landscape similarity, retrieval, and machine mapping of physiographic units | |
CN101551823B (en) | Comprehensive multi-feature image retrieval method | |
CN101763429B (en) | Image retrieval method based on color and shape features | |
CN103577475B (en) | A kind of picture mechanized classification method, image processing method and its device | |
CN102426583A (en) | Chinese medicine tongue manifestation retrieval method based on image content analysis | |
CN108595636A (en) | The image search method of cartographical sketching based on depth cross-module state correlation study | |
CN107679250A (en) | A kind of multitask layered image search method based on depth own coding convolutional neural networks | |
CN103218619A (en) | Image aesthetics evaluating method | |
CN104778458B (en) | A kind of textile pattern search method based on textural characteristics | |
CN101706780A (en) | Image semantic retrieving method based on visual attention model | |
CN102176208A (en) | Robust video fingerprint method based on three-dimensional space-time characteristics | |
CN106446004A (en) | Digital pathological full-slice image retrieval method | |
CN105160317A (en) | Pedestrian gender identification method based on regional blocks | |
CN105574063A (en) | Image retrieval method based on visual saliency | |
CN103714349A (en) | Image recognition method based on color and texture features | |
CN104462526B (en) | More people towards high-resolution remote sensing image cooperate with rapid vector method online | |
CN103345760B (en) | A kind of automatic generation method of medical image object shapes template mark point | |
CN103761295A (en) | Automatic picture classification based customized feature extraction algorithm for art pictures | |
CN106372624A (en) | Human face recognition method and human face recognition system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C12 | Rejection of a patent application after its publication | ||
RJ01 | Rejection of invention patent application after publication |
Open date: 20100203 |