CN101295317B - Method and device for establishing image database index and image recognition - Google Patents
Method and device for establishing image database index and image recognition Download PDFInfo
- Publication number
- CN101295317B CN101295317B CN2008101153278A CN200810115327A CN101295317B CN 101295317 B CN101295317 B CN 101295317B CN 2008101153278 A CN2008101153278 A CN 2008101153278A CN 200810115327 A CN200810115327 A CN 200810115327A CN 101295317 B CN101295317 B CN 101295317B
- Authority
- CN
- China
- Prior art keywords
- color
- image
- branch
- node
- images
- 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
- 238000000034 method Methods 0.000 title claims abstract description 65
- 239000003086 colorant Substances 0.000 claims abstract description 20
- 238000004040 coloring Methods 0.000 claims description 8
- 239000000203 mixture Substances 0.000 description 5
- 230000008878 coupling Effects 0.000 description 4
- 238000010168 coupling process Methods 0.000 description 4
- 238000005859 coupling reaction Methods 0.000 description 4
- 238000003909 pattern recognition Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 239000000284 extract Substances 0.000 description 2
- 235000017060 Arachis glabrata Nutrition 0.000 description 1
- 241001553178 Arachis glabrata Species 0.000 description 1
- 235000010777 Arachis hypogaea Nutrition 0.000 description 1
- 235000018262 Arachis monticola Nutrition 0.000 description 1
- 235000020232 peanut Nutrition 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a method for establishing an index of an image database, comprising the steps that: a set containing all images is set as the current set; non-transparent regions with fully overlapped positions and mutually different colors in all the images of the current set are determined, and the position coordinate and the color of at least one pixel in the non-transparent region of each image is determined as the index of the image. The invention further discloses a device for establishing the index of the image database, comprising a set initialization module and an index establishment module. The method and device for establishing the index of the image database provided by the embodiment of the invention can reduce the occupied storage resources of the image database. The invention simultaneously discloses an image recognition method and device, which can rapidly position the image to be recognized in the image database with a tree structure by the position coordinate and the color information of the position coordinate contained in the index of the database, thus greatly improving the image recognition speed.
Description
Technical field
The present invention relates to image processing techniques, be specifically related to set up the method and apparatus of image data base, also relate to the method and apparatus of image recognition.
Background technology
Along with the current computer technology rapid development, computer process ability is greatly improved; Meanwhile, the correlation technique in field such as pattern-recognition, Flame Image Process has also obtained fast development.
The identification of TV station station symbol, the technology that is meant the utilization Flame Image Process extracts the TV station's station symbol image in the television image, and the process that this station symbol image is discerned.
At present, TV station's station symbol means of identification of main flow adopts the method for pattern-recognition usually in the field of image recognition: promptly in advance the view data of various station symbols is carried out templating and handle, the combination that Templated station symbol data and described station symbol is arranged in information such as the coordinate position of television image, wide high size, color and TV station's title is saved in the characteristic information storehouse as unique characteristic information group corresponding to some station symbols; When carrying out station symbol identification, from described characteristic information storehouse, take out the characteristic information of each station symbol by group; According to data messages such as the coordinate position that writes down in the characteristic information, wide high size and colors, from current television image, gather image to be identified, station symbol template in described image to be identified and this stack features information is carried out pattern match, draw a match-percentage result; With all station symbol templates of described image to be identified and preservation execution pattern coupling and obtain corresponding match-percentage result respectively all, whether a highest match-percentage result is higher than pre-set threshold (be also referred to as coupling confidence threshold value or fiducial probability usually, for example can get 80%) in relatively more all then matching results.If the pairing station symbol of described template that then will have the highest match-percentage result is as the station symbol in the image to be identified.
Use the mode of above-mentioned pattern-recognition can realize the identification of station symbol, but this mode need be preserved the mass data that is comprised in the station symbol image in station symbol template and the image to be identified, therefore the station symbol image data base of being set up is bulky; Simultaneously, in the process of carrying out pattern-recognition to the data of preserving carry out item by item coupling and during comparison because the quantity of station symbol template is not to be a peanut usually, so entire identification process usually can expend considerable time and just can obtain net result.
Summary of the invention
The embodiment of the invention provides a kind of method and apparatus of setting up image data base, can reduce the storage resources that image data base takies.
The embodiment of the invention also provides a kind of method and apparatus of image recognition, can realize image recognition fast with lower cost.
Be first aspect that achieves the above object, technical scheme of the present invention specifically is achieved in that
A kind of method of setting up the image data base index, this method comprises:
The set that a, setting comprise all images is current set;
B, determine under the current set the overlapping fully and nontransparent zone that there are differences of color each other, position in all images, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images.
Described step b comprises:
From nontransparent zone, select a position coordinates as effective coordinate node, and generating the color branch of this effective coordinate node in the color of the pixel of this position coordinates according to all images under the current set, each color branch is included in all images that position coordinates place that this effective coordinate node represents has this color pixel point respectively;
Wherein, only comprise the color branch of an image, the described effective coordinate node in color branch place of an image and all superior nodes of this effective coordinate node of only comprising, be registered as the index of the included image of the described color branch that only comprises an image.
Before described step b, this method further comprises: b0, judge the whether overlapping fully and nontransparent zone that there are differences of color each other, location of all images under the current set; If, then carry out described step b, otherwise, the sky node being set and generating the traversal branch of this sky node, all images that comprises with each traversal branch is current set and continues to carry out this step respectively;
And, described according to current set under all images after the color of the pixel of this position coordinates generates the color branch of this effective coordinate node, described step b further comprises: judge whether all colours branch all only comprises an image, if, process ends then, otherwise all images that comprises with each color branch is current set and returns step b0 respectively.
Describedly from nontransparent zone, select a position coordinates, and comprise at the color branch that the color of the pixel of this position coordinates generates this effective coordinate node according to all images under the current set as effective coordinate node:
With position in all images fully overlapping and each other the whole pixels in the nontransparent zone that there are differences of color as a set T;
The position coordinates of each pixel that comprises among the T with set is as both candidate nodes, calculate under the current set all images respectively in the quantity of each both candidate nodes place color, the both candidate nodes of the minimum number of colors of correspondence as final effective coordinate node, and is respectively the color branch that all images that has same color at this effective coordinate node place under the current set generates this color correspondence.
The described traversal branch that the sky node is set and generates this sky node is:
The sky node is set,, calculates and to have the image number that comprises in the set that opaque pixel is overlapping fully and the image part that there are differences of color each other is formed, generate the traversal branch for comprising the maximum set of image number to each image under the current set; Again image remaining in the current set is repeated described calculating and have the image number that comprises in the set that opaque pixel is overlapping fully and the image part that there are differences of color each other is formed, for comprising the step that the maximum set of image number generates the traversal branch, all images under current set all is included into corresponding traversal branch.
A kind of device of setting up the image data base index, this device comprises: module set up in set initialization module and index;
Described set initialization module, being used to be provided with the set that comprises all images is current set;
Module set up in described index, be used for determining the overlapping fully and nontransparent zone that there are differences of color each other, all images position under the current set, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images.
Described index is set up and is comprised color branch generation unit and index record unit in the module;
Described color branch generation unit, be used for selecting a position coordinates as effective coordinate node from nontransparent zone, and generate the color branch of this effective coordinate node in the color of the pixel of this position coordinates according to all images under the current set, each color branch is included in all images that position coordinates place that this effective coordinate node represents has this color pixel point respectively, notifies the index record unit then;
Described index record unit, be used for only to comprise the color branch of an image, described color branch and effective coordinate node at place and all superior nodes of this effective coordinate node that only comprise an image are recorded as the index of the image that the described color branch that only comprises an image comprised.
The described device of setting up the image data base index further comprises the region decision module, and described index is set up and further comprised traversal branch generation unit in the module,
Described region decision module is used to judge the whether overlapping fully and nontransparent zone that there are differences of color each other, location of all images under the current set; If then with judged result and described non-transparent area field notice color branch generation unit; Otherwise notice traversal branch generation unit;
Described index record unit, be further used for judging whether all colours branch all only comprises an image, if then finish the foundation of database index, otherwise all images that comprises with each color branch is current set and notifies described region decision module respectively;
Described traversal branch generation unit is used to the traversal branch that the sky node is set and generates this sky node, and all images that comprises with each traversal branch is current set and notifies described region decision module respectively.
As seen from the above technical solutions, this method and apparatus of setting up the image data base index of the embodiment of the invention, select the part pixel as the index of each image in image data base by position in all images in the fully overlapping and nontransparent zone that color there are differences each other, and adopt tree structure to store to described index, avoided all images data of memory image template and image to be identified, thereby the storage resources that the image data base of setting up takies significantly reduces.
Be second aspect that achieves the above object, technical scheme of the present invention specifically is achieved in that
A kind of method of image recognition, this method comprises:
The set that a, setting comprise all images is current set;
B, determine under the current set the overlapping fully and nontransparent zone that there are differences of color each other, position in all images, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images;
C, in image data base, search, the matching image that finds is defined as the recognition result of image to be identified according to the index of image to be identified.
Described step b comprises:
From nontransparent zone, select a position coordinates as effective coordinate node, and generating the color branch of this effective coordinate node in the color of the pixel of this position coordinates according to all images under the current set, each color branch is included in all images that position coordinates place that this effective coordinate node represents has this color pixel point respectively;
Wherein, only comprise the color branch of an image, the described effective coordinate node in color branch place of an image and all superior nodes of this effective coordinate node of only comprising, be registered as the index of the included image of the described color branch that only comprises an image.
Before described step b, this method further comprises: b0, judge the whether overlapping fully and nontransparent zone that there are differences of color each other, location of all images under the current set; If, then carry out described step b, otherwise, the sky node being set and generating the traversal branch of this sky node, all images that comprises with each traversal branch is current set and continues to carry out this step respectively;
And, described according to current set under all images after the color of the pixel of this position coordinates generates the color branch of this effective coordinate node, described step b further comprises: judge whether all colours branch all only comprises an image, if, process ends then, otherwise all images that comprises with each color branch is current set and returns step b0 respectively.
Describedly from nontransparent zone, select a position coordinates, and comprise at the color branch that the color of the pixel of this position coordinates generates this effective coordinate node according to all images under the current set as effective coordinate node:
With position in all images fully overlapping and each other the whole pixels in the nontransparent zone that there are differences of color as a set T;
The position coordinates of each pixel that comprises among the T with set is as both candidate nodes, calculate under the current set all images respectively in the quantity of each both candidate nodes place color, the both candidate nodes of the minimum number of colors of correspondence as final effective coordinate node, and is respectively the color branch that all images that has same color at this effective coordinate node place under the current set generates this color correspondence.
The described traversal branch that the sky node is set and generates this sky node is:
The sky node is set,, calculates and to have the image number that comprises in the set that opaque pixel is overlapping fully and the image part that there are differences of color each other is formed, generate the traversal branch for comprising the maximum set of image number to each image under the current set; Again image remaining in the current set is repeated described calculating and have the image number that comprises in the set that opaque pixel is overlapping fully and the image part that there are differences of color each other is formed, for comprising the step that the maximum set of image number generates the traversal branch, all images under current set all is included into corresponding traversal branch.
Described step c is:
C1, obtain image to be identified, the inquiry pointer is set, and the present node that its sensing is set is the start node of image data base, then execution in step c2;
C2, judge whether present node is the sky node, if the node that an optional not traversal branch of inquiry that should the sky node comprises is provided with its sensing is a present node, returns execution in step c2; Otherwise,, obtain colouring information from the corresponding position of image to be identified, then execution in step c3 according to the position coordinates of present node correspondence;
C3, judge the color branch of the color-match that whether exists under the present node and obtain:, then when present node belongs to certain traversal branch, return execution in step c2 if do not exist; If exist, then when comprising a plurality of image, described color branch returns execution in step c2, and when described color branch only comprised an image, this image was the recognition result of image to be identified.
Comprise among the step c3: under described present node, do not have the color branch with the color-match obtained, and present node is judged not comprise image to be identified in the described image data base when not belonging to any one traversal branch.
A kind of device of image recognition, this device comprises: database is set up module and picture recognition module;
Described database is set up module, and being used to be provided with the set that comprises all images is current set; Determine under the current set the overlapping fully and nontransparent zone that there are differences of color each other, position in all images, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images;
Described picture recognition module is used for searching in image data base according to the index of image to be identified, the matching image that finds is defined as the recognition result of image to be identified.
Described database is set up module and comprised: submodule set up in set initialization submodule and index;
Described set initialization submodule, being used to be provided with the set that comprises all images is current set;
Submodule set up in described index, be used for determining the overlapping fully and nontransparent zone that there are differences of color each other, all images position under the current set, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images.
Described index is set up and is comprised color branch generation unit and index record unit in the submodule;
Described color branch generation unit, be used for selecting a position coordinates as effective coordinate node from nontransparent zone, and generate the color branch of this effective coordinate node in the color of the pixel of this position coordinates according to all images under the current set, each color branch is included in all images that position coordinates place that this effective coordinate node represents has this color pixel point respectively, notifies the index record unit then;
Described index record unit, be used for only to comprise the color branch of an image, described color branch and effective coordinate node at place and all superior nodes of this effective coordinate node that only comprise an image are recorded as the index of the image that the described color branch that only comprises an image comprised.
Described database is set up and is further comprised the region decision submodule in the module, and described index is set up and further comprised traversal branch generation unit in the submodule,
Described region decision submodule is used to judge the whether overlapping fully and nontransparent zone that there are differences of color each other, location of all images under the current set; If then with judged result and described non-transparent area field notice color branch generation unit; Otherwise notice traversal branch generation unit;
Described index record unit, be further used for judging whether all colours branch all only comprises an image, if then finish the foundation of database index, otherwise all images that comprises with each color branch is current set and notifies described region decision submodule respectively;
Described traversal branch generation unit is used to the traversal branch that the sky node is set and generates this sky node, and all images that comprises with each traversal branch is current set and notifies described region decision unit respectively.
Described picture recognition module comprises: pre-service submodule, node are judged submodule and color branch judgement submodule;
Described pre-service submodule, be used to obtain image to be identified, set up the database index that obtains in the module according to database, it is the start node of image data base that the present node of inquiring about pointer and its sensing being set is set, and the notice node that is provided with of present node is judged submodule;
Described node is judged submodule, is used to judge whether present node is the sky node, if the not traversal branch node pointed of inquiry that one by one should the sky node comprises is made as present node and continues judges whether described each node is the sky node; Otherwise, according to the position coordinates of present node correspondence, obtain colouring information from the corresponding position of image to be identified after notice color branch judge submodule;
Described color branch is judged submodule, is used to judge the color branch of the color-match that whether exists under the present node and obtain: if do not exist, then notify described node to judge submodule when present node belongs to certain traversal branch; If exist, then when comprising a plurality of image, described color branch notify described node to judge submodule, when described color branch only comprises an image, obtain the recognition result that this image is an image to be identified.
As seen from the above technical solutions, the method and apparatus of this image recognition of the embodiment of the invention, by in all images, selecting the part pixel as the index of each image in image data base in the fully overlapping and nontransparent zone that color there are differences each other, position in advance, and adopt tree structure to store to described index, thereby when treating recognition image and discern, needn't treat recognition image and known image template and carry out comprehensive pixel comparison, and colouring information that can be by the position coordinates that comprises in the database index and this position coordinates place location image to be identified fast in the image data base of tree structure can greatly improve the speed of image recognition.
Description of drawings
Fig. 1 is for setting up the schematic flow sheet of the method for image data base in the embodiment of the invention.
Fig. 2 is the composition structural representation of a characteristic sequence tree in the embodiment of the invention.
Fig. 3 is for setting up the schematic flow sheet of the concrete example of image data base in the embodiment of the invention.
Fig. 4 is for setting up the composition structural representation of the device of image data base in the embodiment of the invention.
Fig. 5 is the schematic flow sheet of the method for image recognition in the embodiment of the invention.
Fig. 6 is the composition structural representation of the device of image recognition in the embodiment of the invention.
Embodiment
For making purpose of the present invention, technical scheme and advantage clearer, below with reference to the accompanying drawing embodiment that develops simultaneously, the present invention is described in more detail.
At first, TV station's station symbol has following basic characteristics:
1, TV station's station symbol is present in the middle of the television image all the time;
2, the station symbol of each TV station has uniqueness, and the station symbol of different TV stations has nothing in common with each other;
3, each TV station's station symbol position in television image remains constant;
4, any variation does not take place in TV station's station symbol along with the change of television image, so the color value of opaque pixel remains constant in the station symbol.
These characteristics according to station symbol, we can find, for TV station's station symbol that all need be discerned, on some coordinate positions in television image, the station symbol of each tame TV station all exists opaque pixel to be positioned on these coordinate positions, and the color of these opaque pixels also is not quite similar.Therefore, the pixel that utilizes these coordinate positions to demarcate just can be distinguished the station symbol of each tame TV station.
In the ideal case, if on some coordinate positions, the station symbol of each tame TV station all has the opaques vegetarian refreshments of different colours separately, then obviously this coordinate is just detected and can once whole station symbols be made a distinction; Yet under actual conditions, might not there be this ideal situation, therefore the situation that more may exist is: on some coordinate positions, the station symbol of part TV station has a kind of color, the station symbol of another part TV station then is another color, certainly, also may also have more complicated situation, station symbol such as some TV station is the third color, and is perhaps more.In this case, only on a coordinate position, whole station symbols can't be made a distinction, therefore need utilize more coordinate position that the above-mentioned station symbol that has passed through preliminary classification is further segmented, understand easily, by continuous repetition said method, finally just whole station symbols can be made a distinction mutually.
Therefore, the present invention introduces the notion of station symbol characteristic sequence tree, and described characteristic sequence tree has following characteristics:
1) station symbol characteristic sequence tree is with the root node of coordinate position as tree or subtree, and described root node is called effective coordinate node; With the branch of color of pixel type, be called the color branch as tree; As leaf, be called terminal node with station symbol;
If certain node A will be called the branch that belongs to the A node or the branch under the A node by the branch that this A node sends; And the node that described branch is pointed (for example send two branches by the A node, point to Node B and C) is called the downstream site of A node; Correspondingly, the A node just is called the superior node of Node B, C.
2) be positioned at the station symbol characteristic sequence and set pixel on the represented coordinate position of effective coordinate node, all belong to the opaque pixel on this station symbol image.Because the color of transparent pixels can change along with the variation of television image, opaque pixel then remains constant, therefore has only by checking opaque pixel, could guarantee other correctness of station identification.
When 3) determining effective coordinate node in the set of current station symbol, its optimal strategy is the minimum number of the color branch that makes that this node had.Can reduce the comparison number of times in the tree data structure search procedure in this way effectively, improve the speed of identification.
4) in current station symbol set, the station symbol that on same effective coordinate node, all has opaque pixel, the color branch by different is assigned under the different subtrees.Otherwise, any two terminal nodes if arrive in the path of whole station symbol characteristic sequence tree root node at them, are owned one section identical color branch together, illustrate that then they on the location of pixels of effective coordinate node representative of this section color branch, have identical color.
5) on station symbol characteristic sequence tree, also exist a kind of special node and branch---empty node and traversal branch.If in current station symbol set, without any a coordinate, make all station symbols all have opaque pixel to be positioned on it or color value not identical entirely, then generate an empty node for this station symbol set, this node does not comprise the coordinate figure of pixel, the branch that extracts from this node also no longer is the color branch, but the traversal branch.The traversal branch represents to adopt the mode of traversal to search traversal branch subtrees pointed all under the sky node, finds the station symbol node up to the mode with coupling in a certain subtree.
Understand easily simultaneously, though above all be to be described with regard to the station symbol image, because the similarity of station symbol image and other image, the principle of above-mentioned characteristic sequence tree can be applied to equally to various treatment of picture.
According to the above description, the embodiment of the invention provides a kind of method of setting up image data base, this method flow as shown in Figure 1, comprising:
Step 101: it is current set that the set that comprises all images is set;
Step 102: determine under the current set the overlapping fully and nontransparent zone that there are differences of color each other, position in all images, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images.
As seen, because position coordinates and color in the index of each image are not identical entirely with other all images, thereby can discern this image uniquely according to the index of each image.
Need to prove that color each other as herein described there are differences and is meant: the complete overlapping nontransparent zone in all images position, each image is incomplete same each other in the color in this nontransparent zone.
For example: in described nontransparent zone, the color of a part of image is a black, and another part is red; Perhaps the color of a part of image is a black, and other images are not a variety of other colors of black etc.
Wherein, step 102 specifically comprises:
From nontransparent zone, select a position coordinates as effective coordinate node, and generating the color branch of this effective coordinate node in the color of the pixel of this position coordinates according to all images under the current set, each color branch is included in all images that position coordinates place that this effective coordinate node represents has this color pixel point respectively;
Wherein, only comprise color branch, this effective coordinate node in color branch place of an image, all superior nodes of this effective coordinate node, be registered as the index of the included image of this color branch.
Before described step 102, this method further comprises:
102a, judge the whether overlapping fully and nontransparent zone that there are differences of color each other, location of all images under the current set; If, then carry out described step 102, otherwise, the sky node being set and generating the traversal branch of this sky node, all images that comprises with each traversal branch is current set and continues to carry out this step respectively;
And, described according to current set under all images after the color of the pixel of this position coordinates generates the color branch of this effective coordinate node, described step 102 further comprises: judge whether all colours branch all only comprises an image, if, process ends then, otherwise all images that comprises with each color branch is current set and returns step 102a respectively.
Preferably, minimum number for the color branch that makes generation, describedly from nontransparent zone, select a position coordinates, and comprise at the color branch that the color of the pixel of this position coordinates generates this effective coordinate node according to all images under the current set as effective coordinate node:
With opaque pixel fully overlapping and each other the whole pixels in the part that there are differences of color as a set T;
The position coordinates of each pixel that comprises among the T with set is as both candidate nodes, calculate under the current set all images respectively in the quantity of each both candidate nodes place color, the both candidate nodes of the minimum number of colors of correspondence as final effective coordinate node, and is respectively the color branch that all images that has same color at this effective coordinate node place under the current set generates this color correspondence.
Preferably, for the minimum number of the traversal branch that makes generation, described the sky node is set and generate this sky node the traversal branch can for:
To each image under the current set, calculate and to have the image number that comprises in the set that opaque pixel is overlapping fully and the image part that there are differences of color each other is formed, generate the traversal branch for comprising the maximum set of image number; Again image remaining in the current set is repeated described calculating and have the image number that comprises in the set that opaque pixel is overlapping fully and the image part that there are differences of color each other is formed, for comprising the step that the maximum set of image number generates the traversal branch, all images under current set all is included into corresponding traversal branch.
Fig. 2 shows the concrete composition structure of a characteristic sequence tree, and wherein (Xi Yi) is effective coordinate node i=1,2 ... 6, each effective node is color branch (arrow shown in the solid line) to the branch that downstream site extends; And the branch that each empty node (comprising sky node 1 and 2) extends to downstream site is traversal branch (arrow shown in the dotted line); Station symbol j (j=1,2 ... 9) be terminal node, station symbol of each terminal node unique identification.
Structure below in conjunction with the characteristic sequence shown in Fig. 2 tree further specifies the method for setting up image data base provided by the invention, flow process as shown in Figure 3, comprising:
Step 300: collect the station symbol data of each channel, described station symbol data comprise the channel information of position coordinates, color and the representative of opaques vegetarian refreshments in television image in each station symbol, continue execution in step 310;
Step 310: whether the opaque pixel coordinate of judging all station symbols in the current set exists overlapping fully part and color incomplete same, if, execution in step 311, otherwise execution in step 316;
Step 311: pixel coordinate that will be fully overlapping is as a set T, gets any pixel coordinate among the set T as both candidate nodes, continues execution in step 312;
Step 312: calculate the quantity of using the color branch that generates under the situation of this both candidate nodes as root node, continue execution in step 313;
Step 313: judge whether to have traveled through all pixel coordinates among the set T, if, execution in step 315, otherwise execution in step 314;
Step 314: it is both candidate nodes that next coordinate pixel is set, and returns step 312;
Step 315: the both candidate nodes of selecting to comprise color branch minimum number generates the color branch and continues execution in step 320 according to this node as final effective coordinate node;
Step 316: generate empty node, continue execution in step 317;
Step 317: under current set, the unappropriated station symbol, seek the maximum station symbol subclass of composition that opaque pixel coordinate is overlapping fully and the lap color is incomplete same, continue execution in step 318;
Step 318:, continue execution in step 319 for selected station symbol subclass generates the traversal branch;
Step 319: judge whether the traversal branch that generates has comprised the whole station symbols under the current set, if, execution in step 320, otherwise return step 317;
Step 320: travel through the whole branches under the current set, continue execution in step 321;
Step 321: judge whether only comprise a station symbol under this branch, if, execution in step 323, otherwise execution in step 322;
Step 322: the set of setting all station symbols that comprise under the current branch is current set, returns step 310;
Step 323: generate the terminal node of this station symbol, continue execution in step 324;
Step 324: judge whether to have traveled through the whole branches under the current set, if, execution in step 325, otherwise return step 320;
Step 325: judge whether current branch place node is the root node of characteristic sequence tree, if, execution in step 327, otherwise execution in step 326;
Step 326: return superior node, the set that all station symbols that comprise under the branch that is untreated of superior node are set is current set, returns step 310;
Step 327: process ends, finish the foundation of image data base.
The embodiment of the invention also provides a kind of device of setting up image data base, and it forms structure as shown in Figure 4, comprising: module 420 set up in set initialization module 410 and index;
Described set initialization module 410, being used to be provided with the set that comprises all images is current set;
Described index is set up and is comprised color branch generation unit 421 and index record unit 422 in the module 420;
Described color branch generation unit 421, be used for selecting a position coordinates as effective coordinate node from nontransparent zone, and generate the color branch of this effective coordinate node in the color of the pixel of this position coordinates according to all images under the current set, each color branch is included in all images that position coordinates place that this effective coordinate node represents has this color pixel point respectively, notifies index record unit 422 then;
Described index record unit 422 is used for and will only comprises the color branch of an image, and effective coordinate node at this color branch and place thereof, all superior nodes of this effective coordinate node are recorded as the index of the image that this color branch comprised;
Further comprise region decision module 430 in the described device of setting up image data base, described index is set up and is further comprised traversal branch generation unit 423 in the module 420;
Described region decision module 430 is used to judge the whether overlapping fully and nontransparent zone that there are differences of color each other, location of all images under the current set; If then with judged result and described non-transparent area field notice color branch generation unit 421; Otherwise notice traversal branch generation unit 423;
Described index record unit 422, be further used for judging whether all colours branch all only comprises an image, if then finish the foundation of database index, otherwise all images that comprises with each color branch is current set and notifies described region decision module 430 respectively;
Described traversal branch generation unit 423 is used to the traversal branch that the sky node is set and generates this sky node, and all images that comprises with each traversal branch is current set and notifies described region decision module 430 respectively.
By as seen above-mentioned, the method and apparatus of setting up the image data base index that the embodiment of the invention provides, select the part pixel as the index of each image in image data base by position in all images in the fully overlapping and nontransparent zone that color there are differences each other, and adopt tree structure to store to described index, avoided all images data of memory image template and image to be identified, thereby the storage resources that the image data base of setting up takies significantly reduces.
According to the image data base that described scheme is set up, the embodiment of the invention further proposes a kind of method of carrying out image recognition based on described image data base, comprising:
Step 1: it is current set that the set that comprises all images is set;
Step 2: determine under the current set the overlapping fully and nontransparent zone that there are differences of color each other, position in all images, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images;
Step 3: the index according to image to be identified is searched in image data base, the matching image that finds is defined as the recognition result of image to be identified.
Wherein, step 1 and step 2 are to set up the step of image data base index, and step 3 is the step of image recognition.The described step of setting up the image data base index is identical with method shown in Figure 1, so locate to repeat no more.
And the flow process of step 3 is as shown in Figure 5, specifically comprises:
Preferably, in step 503, further comprise: under described present node, do not have the color branch with the color-match obtained, and present node is judged not comprise image to be identified in the described image data base when not belonging to any one traversal branch.
Adopt the recognition methods under the prior art, suppose to comprise 500 pixels in the middle of the image to be identified, preserve the template of 100 known image in the database, then each comparison identification all needs the pixel through 50000 times more just can obtain net result; And even the method for the employing image recognition that the embodiment of the invention provided in the worst case, also only needs 100 pixels relatively, can obtain recognition result.As seen, the method for the image recognition that the embodiment of the invention provided remains 500 times of existing recognition speed under the slowest situation of speed.
The embodiment of the invention also provides a kind of device of image recognition, forms structure as shown in Figure 6, comprising: database is set up module 610 and picture recognition module 620;
Described database is set up module 610, and being used to be provided with the set that comprises all images is current set; Determine under the current set the overlapping fully and nontransparent zone that there are differences of color each other, position in all images, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images;
Described picture recognition module 620 is used for searching in image data base according to the index of image to be identified, the matching image that finds is defined as the recognition result of image to be identified.
Described database is set up module 610 and comprised: submodule 612 set up in set initialization submodule 611 and index;
Described set initialization submodule 611, being used to be provided with the set that comprises all images is current set;
Submodule 612 set up in described index, be used for determining the overlapping fully and nontransparent zone that there are differences of color each other, all images position under the current set, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images.
Described index is set up and is comprised color branch generation unit 614 and index record unit 615 in the submodule 612;
Described color branch generation unit 614, be used for selecting a position coordinates as effective coordinate node from nontransparent zone, and generate the color branch of this effective coordinate node in the color of the pixel of this position coordinates according to all images under the current set, each color branch is included in all images that position coordinates place that this effective coordinate node represents has this color pixel point respectively, notifies index record unit 615 then;
Described index record unit 615 is used for and will only comprises the color branch of an image, and effective coordinate node at this color branch and place thereof, all superior nodes of this effective coordinate node are recorded as the index of the image that this color branch comprised.
Described database is set up and is further comprised region decision submodule 613 in the module 610, and described index is set up and further comprised traversal branch generation unit 616 in the submodule 612;
Described region decision submodule 613 is used to judge the whether overlapping fully and nontransparent zone that there are differences of color each other, location of all images under the current set; If then with judged result and described non-transparent area field notice color branch generation unit 614; Otherwise notice traversal branch generation unit 616;
Described index record unit 615, be further used for judging whether all colours branch all only comprises an image, if then finish the foundation of database index, otherwise all images that comprises with each color branch is current set and notifies described region decision module 613 respectively;
Described traversal branch generation unit 616 is used to the traversal branch that the sky node is set and generates this sky node, and all images that comprises with each traversal branch is current set and notifies described region decision unit 613 respectively.
Described picture recognition module 620 comprises: pre-service submodule 621, node are judged submodule 622 and color branch judgement submodule 623;
Described pre-service submodule 621, be used to obtain image to be identified, set up the database index that obtains in the module 610 according to database, it is the start node of database index that the present node of inquiring about pointer and its sensing being set is set, and the notice node that is provided with of present node is judged submodule 622;
Described node is judged submodule 622, is used to judge whether present node is the sky node, if the not traversal branch node pointed of inquiry that one by one should the sky node comprises is made as present node and continues judges whether described each node is the sky node; Otherwise, according to the position coordinates of present node correspondence, obtain colouring information from the corresponding position of image to be identified after notice color branch judge submodule 623;
Described color branch is judged submodule 623, is used to judge the color branch of the color-match that whether exists under the present node and obtain: if do not exist, then notify described node to judge submodule 622 when present node belongs to certain traversal branch; If exist, then when comprising a plurality of image, described color branch notify described node to judge submodule 622, when described color branch only comprises an image, obtain the recognition result that this image is an image to be identified.
By above-mentioned explanation as seen, the method of the image recognition that the embodiment of the invention provides, by in all images, selecting the part pixel as the index of each image in image data base in the fully overlapping and nontransparent zone that color there are differences each other, position in advance, and adopt tree structure to store to described index, thereby when treating recognition image and discern, needn't treat recognition image and known image template and carry out comprehensive pixel comparison, and colouring information that can be by the position coordinates that comprises in the database index and this position coordinates place location image to be identified fast in the image data base of tree structure can greatly improve the speed of image recognition.
The above is preferred embodiment of the present invention only, is not to be used to limit spirit of the present invention and protection domain, and equivalent variations that any those of ordinary skill in the art made or replacement all should be considered as being encompassed within protection scope of the present invention.
Claims (20)
1. a method of setting up the image data base index is characterized in that, this method comprises:
The set that a, setting comprise all images is current set;
B, determine under the current set the overlapping fully and nontransparent zone that there are differences of color each other, position in all images, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images.
2. the method for claim 1 is characterized in that, described step b comprises:
From nontransparent zone, select a position coordinates as effective coordinate node, and generating the color branch of this effective coordinate node in the color of the pixel of this position coordinates according to all images under the current set, each color branch is included in all images that position coordinates place that this effective coordinate node represents has this color pixel point respectively;
Wherein, only comprise the color branch of an image, the described effective coordinate node in color branch place of an image and all superior nodes of this effective coordinate node of only comprising, be registered as the index of the included image of the described color branch that only comprises an image.
3. method according to claim 2 is characterized in that, before described step b, this method further comprises: b0, judge the whether overlapping fully and nontransparent zone that there are differences of color each other, location of all images under the current set; If, then carry out described step b, otherwise, the sky node being set and generating the traversal branch of this sky node, all images that comprises with each traversal branch is current set and continues to carry out this step respectively;
And, described according to current set under all images after the color of the pixel of this position coordinates generates the color branch of this effective coordinate node, described step b further comprises: judge whether all colours branch all only comprises an image, if, process ends then, otherwise all images that comprises with each color branch is current set and returns step b0 respectively.
4. method according to claim 3, it is characterized in that, describedly from nontransparent zone, select a position coordinates, and comprise at the color branch that the color of the pixel of this position coordinates generates this effective coordinate node according to all images under the current set as effective coordinate node:
With position in all images fully overlapping and each other the whole pixels in the nontransparent zone that there are differences of color as a set T;
The position coordinates of each pixel that comprises among the T with set is as both candidate nodes, calculate under the current set all images respectively in the quantity of each both candidate nodes place color, the both candidate nodes of the minimum number of colors of correspondence as final effective coordinate node, and is respectively the color branch that all images that has same color at this effective coordinate node place under the current set generates this color correspondence.
5. method according to claim 4 is characterized in that, the described traversal branch that the sky node is set and generates this sky node is:
The sky node is set,, calculates and to have the image number that comprises in the set that opaque pixel is overlapping fully and the image part that there are differences of color each other is formed, generate the traversal branch for comprising the maximum set of image number to each image under the current set; Again image remaining in the current set is repeated described calculating and have the image number that comprises in the set that opaque pixel is overlapping fully and the image part that there are differences of color each other is formed, for comprising the step that the maximum set of image number generates the traversal branch, all images under current set all is included into corresponding traversal branch.
6. a device of setting up the image data base index is characterized in that, this device comprises: module set up in set initialization module and index;
Described set initialization module, being used to be provided with the set that comprises all images is current set;
Module set up in described index, be used for determining the overlapping fully and nontransparent zone that there are differences of color each other, all images position under the current set, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images.
7. device according to claim 6 is characterized in that, described index is set up and comprised color branch generation unit and index record unit in the module;
Described color branch generation unit, be used for selecting a position coordinates as effective coordinate node from nontransparent zone, and generate the color branch of this effective coordinate node in the color of the pixel of this position coordinates according to all images under the current set, each color branch is included in all images that position coordinates place that this effective coordinate node represents has this color pixel point respectively, notifies the index record unit then;
Described index record unit, be used for only to comprise the color branch of an image, described color branch and effective coordinate node at place and all superior nodes of this effective coordinate node that only comprise an image are recorded as the index of the image that the described color branch that only comprises an image comprised.
8. device according to claim 7 is characterized in that this device further comprises the region decision module, and described index is set up and further comprised traversal branch generation unit in the module,
Described region decision module is used to judge the whether overlapping fully and nontransparent zone that there are differences of color each other, location of all images under the current set; If then with judged result and described non-transparent area field notice color branch generation unit; Otherwise notice traversal branch generation unit;
Described index record unit, be further used for judging whether all colours branch all only comprises an image, if then finish the foundation of database index, otherwise all images that comprises with each color branch is current set and notifies described region decision module respectively;
Described traversal branch generation unit is used to the traversal branch that the sky node is set and generates this sky node, and all images that comprises with each traversal branch is current set and notifies described region decision module respectively.
9. the method for an image recognition is characterized in that, this method comprises:
The set that a, setting comprise all images is current set;
B, determine under the current set the overlapping fully and nontransparent zone that there are differences of color each other, position in all images, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images;
C, in image data base, search, the matching image that finds is defined as the recognition result of image to be identified according to the index of image to be identified.
10. method according to claim 9 is characterized in that, described step b comprises:
From nontransparent zone, select a position coordinates as effective coordinate node, and generating the color branch of this effective coordinate node in the color of the pixel of this position coordinates according to all images under the current set, each color branch is included in all images that position coordinates place that this effective coordinate node represents has this color pixel point respectively;
Wherein, only comprise the color branch of an image, the described effective coordinate node in color branch place of an image and all superior nodes of this effective coordinate node of only comprising, be registered as the index of the included image of the described color branch that only comprises an image.
11. method according to claim 10 is characterized in that, before described step b, this method further comprises: b0, judge the whether overlapping fully and nontransparent zone that there are differences of color each other, location of all images under the current set; If, then carry out described step b, otherwise, the sky node being set and generating the traversal branch of this sky node, all images that comprises with each traversal branch is current set and continues to carry out this step respectively;
And, described according to current set under all images after the color of the pixel of this position coordinates generates the color branch of this effective coordinate node, described step b further comprises: judge whether all colours branch all only comprises an image, if, process ends then, otherwise all images that comprises with each color branch is current set and returns step b0 respectively.
12. method according to claim 11, it is characterized in that, describedly from nontransparent zone, select a position coordinates, and comprise at the color branch that the color of the pixel of this position coordinates generates this effective coordinate node according to all images under the current set as effective coordinate node:
With position in all images fully overlapping and each other the whole pixels in the nontransparent zone that there are differences of color as a set T;
The position coordinates of each pixel that comprises among the T with set is as both candidate nodes, calculate under the current set all images respectively in the quantity of each both candidate nodes place color, the both candidate nodes of the minimum number of colors of correspondence as final effective coordinate node, and is respectively the color branch that all images that has same color at this effective coordinate node place under the current set generates this color correspondence.
13. method according to claim 12 is characterized in that, the described traversal branch that the sky node is set and generates this sky node is:
The sky node is set,, calculates and to have the image number that comprises in the set that opaque pixel is overlapping fully and the image part that there are differences of color each other is formed, generate the traversal branch for comprising the maximum set of image number to each image under the current set; Again image remaining in the current set is repeated described calculating and have the image number that comprises in the set that opaque pixel is overlapping fully and the image part that there are differences of color each other is formed, for comprising the step that the maximum set of image number generates the traversal branch, all images under current set all is included into corresponding traversal branch.
14. method according to claim 13 is characterized in that, described step c is:
C1, obtain image to be identified, the inquiry pointer is set, and the present node that its sensing is set is the start node of image data base, then execution in step c2;
C2, judge whether present node is the sky node, if the node that an optional not traversal branch of inquiry that should the sky node comprises is provided with its sensing is a present node, returns execution in step c2; Otherwise,, obtain colouring information from the corresponding position of image to be identified, then execution in step c3 according to the position coordinates of present node correspondence;
C3, judge the color branch of the color-match that whether exists under the present node and obtain:, then when present node belongs to certain traversal branch, return execution in step c2 if do not exist; If exist, then when comprising a plurality of image, described color branch returns execution in step c2, and when described color branch only comprised an image, this image was the recognition result of image to be identified.
15. method according to claim 14, it is characterized in that, comprise among the step c3: under described present node, do not have the color branch with the color-match obtained, and present node is judged not comprise image to be identified in the described image data base when not belonging to any one traversal branch.
16. the device of an image recognition is characterized in that, this device comprises: database is set up module and picture recognition module;
Described database is set up module, and being used to be provided with the set that comprises all images is current set; Determine under the current set the overlapping fully and nontransparent zone that there are differences of color each other, position in all images, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images;
Described picture recognition module is used for searching in image data base according to the index of image to be identified, the matching image that finds is defined as the recognition result of image to be identified.
17. device according to claim 16 is characterized in that, described database is set up module and comprised: submodule set up in set initialization submodule and index;
Described set initialization submodule, being used to be provided with the set that comprises all images is current set;
Submodule set up in described index, be used for determining the overlapping fully and nontransparent zone that there are differences of color each other, all images position under the current set, and from the described nontransparent zone of each image, determine the position coordinates of at least one pixel and color index as this image; Wherein, position coordinates in the index of each image and color are not identical entirely with other all images.
18. device according to claim 17 is characterized in that, described index is set up and is comprised color branch generation unit and index record unit in the submodule;
Described color branch generation unit, be used for selecting a position coordinates as effective coordinate node from nontransparent zone, and generate the color branch of this effective coordinate node in the color of the pixel of this position coordinates according to all images under the current set, each color branch is included in all images that position coordinates place that this effective coordinate node represents has this color pixel point respectively, notifies the index record unit then;
Described index record unit, be used for only to comprise the color branch of an image, described color branch and effective coordinate node at place and all superior nodes of this effective coordinate node that only comprise an image are recorded as the index of the image that the described color branch that only comprises an image comprised.
19. device according to claim 18 is characterized in that, described database is set up and is further comprised the region decision submodule in the module, and described index is set up and further comprised traversal branch generation unit in the submodule,
Described region decision submodule is used to judge the whether overlapping fully and nontransparent zone that there are differences of color each other, location of all images under the current set; If then with judged result and described non-transparent area field notice color branch generation unit; Otherwise notice traversal branch generation unit;
Described index record unit, be further used for judging whether all colours branch all only comprises an image, if then finish the foundation of database index, otherwise all images that comprises with each color branch is current set and notifies described region decision submodule respectively;
Described traversal branch generation unit is used to the traversal branch that the sky node is set and generates this sky node, and all images that comprises with each traversal branch is current set and notifies described region decision unit respectively.
20. device according to claim 19 is characterized in that, described picture recognition module comprises: pre-service submodule, node are judged submodule and color branch judgement submodule;
Described pre-service submodule, be used to obtain image to be identified, set up the database index that obtains in the module according to database, it is the start node of image data base that the present node of inquiring about pointer and its sensing being set is set, and the notice node that is provided with of present node is judged submodule;
Described node is judged submodule, is used to judge whether present node is the sky node, if the not traversal branch node pointed of inquiry that one by one should the sky node comprises is made as present node and continues judges whether described each node is the sky node; Otherwise, according to the position coordinates of present node correspondence, obtain colouring information from the corresponding position of image to be identified after notice color branch judge submodule;
Described color branch is judged submodule, is used to judge the color branch of the color-match that whether exists under the present node and obtain: if do not exist, then notify described node to judge submodule when present node belongs to certain traversal branch; If exist, then when comprising a plurality of image, described color branch notify described node to judge submodule, when described color branch only comprises an image, obtain the recognition result that this image is an image to be identified.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008101153278A CN101295317B (en) | 2008-06-20 | 2008-06-20 | Method and device for establishing image database index and image recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008101153278A CN101295317B (en) | 2008-06-20 | 2008-06-20 | Method and device for establishing image database index and image recognition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101295317A CN101295317A (en) | 2008-10-29 |
| CN101295317B true CN101295317B (en) | 2011-03-02 |
Family
ID=40065601
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008101153278A Active CN101295317B (en) | 2008-06-20 | 2008-06-20 | Method and device for establishing image database index and image recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101295317B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102193934B (en) * | 2010-03-11 | 2013-05-29 | 株式会社理光 | System and method for searching representative image of image set |
| CN102662559B (en) * | 2012-03-13 | 2015-04-29 | 西安交通大学 | Design method of virtual touch screen technology based on scene identification |
| CN103002328A (en) * | 2012-11-30 | 2013-03-27 | 北京十分科技有限公司 | Method and device for identifying channels |
| CN110992443B (en) * | 2019-12-18 | 2023-09-26 | 政采云有限公司 | Business flow chart construction method and device and computer readable storage medium |
-
2008
- 2008-06-20 CN CN2008101153278A patent/CN101295317B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN101295317A (en) | 2008-10-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Progressive learning with multi-scale attention network for cross-domain vehicle re-identification | |
| CN101587478B (en) | Methods and devices for training, automatically labeling and searching images | |
| CN101944187B (en) | Two-dimensional micro code and processing method and device thereof | |
| CN103400099B (en) | Terminal and two-dimensional code identification method | |
| CN103049728A (en) | Method, system and terminal for augmenting reality based on two-dimension code | |
| CN101295317B (en) | Method and device for establishing image database index and image recognition | |
| CN102511048B (en) | A kind of method and system comprising the video area of text for pre-service | |
| CN103970733B (en) | A kind of Chinese new word identification method based on graph structure | |
| CN106502891A (en) | Automatic detection method and device for user interface | |
| CN103034831A (en) | Method and system for identifying linear bar code | |
| CN101976333A (en) | Method for automatically distinguishing first-generation identity card from second-generation identity card | |
| CN100447793C (en) | Method for extracting page query interface based on character of vision | |
| CN104376291B (en) | The method and device of data processing | |
| CN101930532A (en) | Mobile phone-based camera quick response matrix code recognizing and reading method | |
| CN101339560B (en) | Method and device for searching series data, and search engine system | |
| CN106933899A (en) | Form processing method and device | |
| CN104063701B (en) | Fast electric television stations TV station symbol recognition system and its implementation based on SURF words trees and template matches | |
| CN105868724A (en) | Two-dimensional bar code identification method and system based on connected domain | |
| Yao et al. | Deep learning for phishing detection | |
| CN105844200A (en) | Method of determining bar code type | |
| Fan et al. | A run-length-coding-based approach to stroke extraction of Chinese characters | |
| CN105160650B (en) | A kind of method extracting the continuous subject image of the degree of depth from image | |
| CN116976375A (en) | Method and device for detecting cigarette packaging box in code scanning and verification scene | |
| JP6365117B2 (en) | Information processing apparatus, image determination method, and program | |
| Arias et al. | Efficient extraction of primitives from line drawings composed of horizontal and vertical lines |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20160519 Address after: 519031 Guangdong city of Zhuhai province Hengqin Baohua Road No. 6, room 105 -478 Patentee after: GUANGDONG ZHONGXING ELECTRONICS CO., LTD. Address before: 100083, Haidian District, Xueyuan Road, Beijing No. 35, Nanjing Ning building, 15 Floor Patentee before: Beijing Vimicro Corporation |