CN203299453U - Electronic telescope system based on image identification - Google Patents
Electronic telescope system based on image identification Download PDFInfo
- Publication number
- CN203299453U CN203299453U CN2013202417763U CN201320241776U CN203299453U CN 203299453 U CN203299453 U CN 203299453U CN 2013202417763 U CN2013202417763 U CN 2013202417763U CN 201320241776 U CN201320241776 U CN 201320241776U CN 203299453 U CN203299453 U CN 203299453U
- Authority
- CN
- China
- Prior art keywords
- image
- lens
- treating apparatus
- processing
- module
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
The utility model discloses an electronic telescope system based on image identification. The electronic telescope system comprises a pedestal, and an electronic telescope body on a support of the pedestal, and the electronic telescope body further comprises a lens barrel installed on the support, and a lens arranged at the front end of the lens barrel, a viewing window at the rear end of the lens barrel, a processing device for acquiring and processing images collected by the lens, and a display screen for receiving and displaying image data after the images being processed by the processing device. By adding a processing device for processing images, the electronic telescope system can conduct complex process on images acquired by the lens and classify the images. The electronic telescope system is simple in structure and operation and is suitable for middle and high-grade electron telescope.
Description
Technical field
The utility model has related to a kind of electron telescope, relates in particular to a kind of system of electron telescope based on image recognition of son.
Background technology
Telescope is a kind of optical instrument that utilizes concavees lens and the remote object of convex lens observation.Utilize light refraction or the light of scioptics are made it to enter aperture and are converged to picture by reflection shielding film, then through a magnification eyepiece and in sight.Telescope mainly contains two effects, and the one, the subtended angle for amplifying distant objects, make human eye can see the less details of angular distance clearly; The 2nd, object lens collect than pupil diameter (maximum 8 millimeters) thick the light beam of Duo, send into human eye, the observer can be seen originally can't see dark a little less than object.
Electron telescope is the perfect crystallization of modern electronic technology and optical technology.According to different purposes, the morphology and function of product is not identical yet.Military electron telescope combines night vision technology; Civilian electron telescope has all incorporated anti-shake technology mostly, make hand-held telescopical user can obtain better using impression, be used for the equatorial telescope that astronomical electron telescope has used electronics, make observation more light and accurate, pass through electronic romote-controller, telescopical position can be controlled accurately, and video recorder can be connected and computer is recorded and takes pictures.
The image effect that gathers for also improving electron telescope, need to process the image after adopting, and traditional electron telescope does not possess image processing function or just simply processes, and does not possess the functions such as complicated image processing and graphic collection.
The utility model content
The technical problems to be solved in the utility model is in order to overcome technological deficiency recited above, and a kind of system of electron telescope based on image recognition is provided, and the image processing method of planting based on the electron telescope system of image recognition also is provided simultaneously.
In order to solve technical matters recited above, the utility model is taked following technical scheme:
A kind of system of electron telescope based on image recognition, include base, be arranged on the electron telescope body on the support of base, described electron telescope body further includes rack-mount lens barrel, be arranged at the lens barrel front end camera lens, be arranged at the Guan Pingkou of lens barrel rear end, also include be used to obtaining the image that camera lens collects and the treating apparatus of processing, be used for the view data after receiving and processing device is processed and the display screen that shows.
Described treating apparatus further includes: be used for to control image capture module that camera lens carries out image acquisition, be used for receiving image information and the picture recognition module of identifying that image capture module collects, be used for preserving the temporary module of data of the image information after picture recognition module identification and the data analysis module that is used for the image information after picture recognition module identification is analyzed and transferred to display screen.
Described treating apparatus is the processing integration module that is built in lens barrel, and described processing integration module is connection lens and display screen simultaneously.
Described treating apparatus is external treating apparatus, and described treating apparatus is computing machine, and described computing machine is connection lens and display screen simultaneously.
Described camera lens is pick-up lens.
A kind of image processing method of the system of the electron telescope based on image recognition, comprise the steps:
The image input step, gather image and the input picture analysis module is processed by image capture module;
The anti-treatment step of image, carry out opposite direction to the image that inputs to image analysis module and process, and makes image and material object after processing consistent;
Characteristic extraction step, analyze the image after processing by data analysis module, the formal representation of characteristics of image use data that can the presentation video uniqueness out;
The Images Classification step, the image that will have the inhomogeneity characteristics of image is assigned in different image data bases.
Before carrying out characteristic extraction step, also comprise the image pre-treatment step, before carrying out feature extraction, the image to after anti-processing carries out pre-service.
Also comprise the images match step, on the basis of image pre-service and feature extraction, the new images feature of current input and the template image feature that is kept in image data base are in advance compared, judge whether two characteristics of image are consistent, if the new images feature is consistent with the template image feature, access the data message corresponding with the new images feature and mark processing from image data base, the target image after being marked, the target image after marking finally is presented on display screen.
Wherein, characteristic extraction step is limited to the quantity of Images Classification can distinguish the image degree that should be recognition result.Above-mentioned image data base comprises Hash table, and the index that this Hash table is calculated according to each local feature vectors that derives from each image with process is in accordance with regulations classified to it.
In the image pre-treatment step, remove the lower proper vector of resolving ability, only register image reference data corresponding to the proper vector higher with resolving ability in Hash table, proper vector that therefore can be only that resolving ability is higher is carried out image recognition at short notice as processing object.In addition, only register image reference data corresponding to the proper vector higher with resolving ability in Hash table, therefore with the situation of registering the image reference data corresponding with all proper vectors, compare, can save the required memory span of image data base.
The image-recognizing method of picture recognition module of the present utility model is the method that the use characteristic vector carrys out recognition image.The contrast of the proper vector that substantially is to be registered in proper vector in database and input picture of identification.The local feature of proper vector presentation video, therefore obtain a plurality of proper vectors usually from an image.But, be registered in the proper vector of the object in database and have the proper vector of the feature that represents well this object and be not the proper vector that represents well this object features.The proper vector that represents well the feature of object refers to if having this proper vector can say that input picture is this object, becomes the proper vector of abundant proof.On the other hand, not that the proper vector that represents well the feature of this object refers to following proper vector: in the image that appears at various objects, though therefore have this proper vector, can not be used in to judge it is which object.The removing of proper vector refers to the latter namely can not be deleted as the proper vector of evidence from dictionary processing.
Electron telescope system and image processing method thereof that the utility model provides, be used for by increase the treating apparatus that image is processed, and can carry out to the image of camera lens collection the intricately processing and carry out the graphic collection function.The utility model is simple in structure, operation is simple and easy, is applicable to the electron telescope of all kinds of medium and high classes.
Description of drawings
Fig. 1 is system architecture schematic diagram of the present utility model.
Fig. 2 is the structured flowchart for the treatment of apparatus of the present utility model.
Fig. 3 is image processing flow chart of the present utility model.
In figure, 10. lens barrel, 11. treating apparatus, 12. are seen screen mouth, 20. display screens, 30. bases, 31. fixed supports, 40. cameras.
Embodiment
See also Fig. 1 and Fig. 2, as shown in the figure, a kind of system of electron telescope based on image recognition, include base, be arranged on the electron telescope body on the support of base, the electron telescope body further includes rack-mount lens barrel, be arranged at the lens barrel front end camera lens, be arranged at the Guan Pingkou of lens barrel rear end, camera lens is pick-up lens, also include be used to obtaining the image that camera lens collects and the treating apparatus of processing, be used for the view data after receiving and processing device is processed and the display screen that shows.Treating apparatus further includes: be used for to control image capture module that camera lens carries out image acquisition, be used for receiving image information and the picture recognition module of identifying that image capture module collects, be used for preserving the temporary module of data of the image information after picture recognition module identification and the data analysis module that is used for the image information after picture recognition module identification is analyzed and transferred to display screen.Treating apparatus is the processing integration module that is built in lens barrel, processes integration module connection lens and display screen simultaneously.Perhaps treating apparatus is external treating apparatus, and treating apparatus is computing machine, and computing machine is connection lens and display screen simultaneously.
Gather image and the image transmitting that collects is arrived and processes device 11 by camera lens, the image of the input for the treatment of apparatus 11 reception cameras 40 is also identified image, the data of identification are edited processing with the later stage, and the image after processing is shown by display screen 20, watch thereby can shield mouth 12 by sight.
A kind of image processing method of the system of the electron telescope based on image recognition, comprise the steps:
The image input step, gather image and the input picture analysis module is processed by image capture module;
The anti-treatment step of image, carry out opposite direction to the image that inputs to image analysis module and process, and makes image and material object after processing consistent;
Characteristic extraction step, analyze the image after processing by data analysis module, the formal representation of characteristics of image use data that can the presentation video uniqueness out;
In the anti-treatment step of image, that the image after instead processing is temporary to the temporary module of data.
The Images Classification step, the image that will have the inhomogeneity characteristics of image is assigned in different image data bases.
Before carrying out characteristic extraction step, also comprise the image pre-treatment step, before carrying out feature extraction, the image to after anti-processing carries out pre-service.
Also comprise the images match step, on the basis of image pre-service and feature extraction, the new images feature of current input and the template image feature that is kept in image data base are in advance compared, judge whether two characteristics of image are consistent, if the new images feature is consistent with the template image feature, access the data message corresponding with the new images feature and mark processing from image data base, the target image after being marked, the target image after marking finally is presented on display screen.
For the complexity that reduces subsequent algorithm with raise the efficiency, the image pre-service is an important step.For example: background separation is with image area and background separation, thereby avoids carrying out feature extraction in the zone that there is no effective information, accelerates the speed of subsequent treatment, improves the precision of image characteristics extraction and coupling; The purpose of figure image intensifying is to improve picture quality, recovers its original structure; The binaryzation of image is to be bianry image with image from greyscale image transitions; Image thinning is that clear but inhomogeneous bianry image is changed into the point and line chart picture that live width is only a pixel.In the image pre-treatment step, remove the lower proper vector of resolving ability, only register image reference data corresponding to the proper vector higher with resolving ability in Hash table, proper vector that therefore can be only that resolving ability is higher is carried out image recognition at short notice as processing object.In addition, only register image reference data corresponding to the proper vector higher with resolving ability in Hash table, therefore with the situation of registering the image reference data corresponding with all proper vectors, compare, can save the required memory span of image data base.
Characteristic extraction step, the formal representation of the feature use numerical value that can fully represent this image uniqueness out; Keep real features, the false feature of filtering as far as possible.The contrast of the proper vector that substantially is to be registered in proper vector in database and input picture of identification.The local feature of proper vector presentation video, therefore obtain a plurality of proper vectors usually from an image.But, be registered in the proper vector of the image in database and have the proper vector of the feature that represents well this object and be not the proper vector that represents well this object features.The proper vector that represents well the feature of object refers to if having this proper vector can say that input picture is this object, becomes the proper vector of abundant proof.On the other hand, not that the proper vector that represents well the feature of this object refers to following proper vector: in the image that appears at various objects, though therefore have this proper vector, can not be used in to judge it is which object.Thereby characteristic extraction step of the present utility model namely can not be deleted the latter as the proper vector of evidence from dictionary.
The Images Classification step, in picture system, the image of input will mate with tens of even thousands of images up to a hundred in database.In order to reduce search time, to reduce the complexity of calculating, the utility model is in the object identification of using the approximate KNN search, and degree of approximation becomes the important parameter for balance discrimination and efficiency.Strengthen being similar to more can cutting down the processing time, but when undue reinforcement is approximate, a lot of proper vectors can not asked arest neighbors, as a result of cause causing identification by mistake.To cause that the degree of approximation of mistake identification can be according to image and difference in one of this problem.Significantly be similar to even exist " simply " image that also can identify, exist on the contrary and significantly be similar to " difficulty " image of identification by mistake., in order by fixing, to be similar to guarantee certain discrimination, need to make degree of approximation coordinate the image of identification difficulty, thereby become the obstacle of raising the efficiency.
Therefore,, as an optimal way of the present utility model, from the viewpoint of " degree of accuracy of identifying required nearest neighbor search can be according to image and difference ", provide the method for processing of cutting down.Namely for image, regulate adaptively the method for degree of approximation.The utility model, by preparing the different a plurality of discriminators of degree of approximation and from degree of approximation, connecting them to weak multistage file by force, can be realized regulating adaptively degree of approximation for image.Thus, can identify at high speed the image that can simply identify at the discriminator that front exponent part utilization significantly is similar to, can come only the image that significantly is similar to None-identified time-consumingly critically to be identified at the approximate weak discriminator of rear exponent part utilization.
In addition, in the time can't distinguishing the image that should be recognition result, can also relax the object degree of limit search in above-mentioned restriction operation process and remove outside the partial descriptions symbol that before is considered as object search and determine, thereby set up the processing of new search object, the object search that determines is carried out the search operation and distinguished operation.Thus,, even in the situation that change degree of approximation, carry out and limit operation, search for operation and distinguish operation multistagely, also can compare the not a halfpenny the worse processing time with the situation with search becomes the partial descriptions symbol one time of object search in each rank and identify.
The method is characterized in that the constructive method of the discriminator that carries out multistageization.In the discriminator on rear rank, only will be similar to the difference that difference causes, the proper vector that does not namely become object in the discriminator than its front rank is made as the object that distance is calculated, and can access following advantage thus: proceed to last rank and also only need situation calculated amount about equally with the discriminator on the last rank of independent use even process.And, when in the degree of periodically relaxing the limit search object, repeating above-mentioned restriction operation, search for operation and distinguishing that operation also can't be distinguished the image that should be recognition result, also can refuse the Search Results about this partial descriptions symbol.Thus, compare and can suppress false recognition rate with situation about not refusing.
in addition, above-mentioned image data base comprises Hash table, this Hash table is classified to it with the index value that process is in accordance with regulations calculated each partial descriptions symbol of deriving from each image, above-mentioned restriction operation considers that characteristic quantity change ground accords with calculating index value by said process according to each partial descriptions of input picture, be made as object search with the index value of calculating with reference to the partial descriptions symbol that above-mentioned Hash table also will belong to such, above-mentioned each partial descriptions symbol use statistical treatment of distinguishing operation definite neighbour to passing through the search operation, this statistical treatment is voted for the image to obtaining this each partial descriptions symbol, above-mentioned Hash table is made as follows: to all kinds of, surpass threshold value in the situation that belong to the quantity of such partial descriptions symbol, remove such partial descriptions symbol from object search.Thus, due in the situation that belong to the quantity of all kinds of partial descriptions symbol and surpass threshold value and remove them make Hash table from object search, therefore the partial descriptions symbol that is considered as object search in limiting operation is limited at the higher partial descriptions symbol of resolving ability, thereby realizes identification efficiently.
In a fairly large number of situation of the local feature vectors of a classification index that belongs to Hash table, can say that these local feature vectors resolving abilities are lower.Namely the meaning is in the situation that according to the local feature vectors of input picture, calculate index and, with reference to Hash table, be registered with a plurality of candidates that belong to such.This local feature vectors does not have much contributions to the candidate item of constriction image recognition., if remove the lower local feature vectors of resolving ability from object search, only with reference to the higher local feature vectors of resolving ability, thereby identify efficiently.
Claims (5)
1. system of the electron telescope based on image recognition, include base, be arranged on the electron telescope body on the support of base, described electron telescope body further includes rack-mount lens barrel, be arranged at the lens barrel front end camera lens, be arranged at the Guan Pingkou of lens barrel rear end, it is characterized in that, also include be used to obtaining the image that camera lens collects and the treating apparatus of processing, be used for the view data after receiving and processing device is processed and the display screen that shows.
2. electron telescope system as claimed in claim 1, it is characterized in that, described treating apparatus further includes: be used for to control image capture module that camera lens carries out image acquisition, be used for receiving image information and the picture recognition module of identifying that image capture module collects, be used for preserving the temporary module of data of the image information after picture recognition module identification and the data analysis module that is used for the image information after picture recognition module identification is analyzed and transferred to display screen.
3. electron telescope system as claimed in claim 1 or 2, is characterized in that, described treating apparatus is the processing integration module that is built in lens barrel, and described processing integration module is connection lens and display screen simultaneously.
4. electron telescope system as claimed in claim 1 or 2, is characterized in that, described treating apparatus is external treating apparatus, and described treating apparatus is computing machine, and described computing machine is connection lens and display screen simultaneously.
5. electron telescope system as claimed in claim 1, it is characterized in that: described camera lens is pick-up lens.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013202417763U CN203299453U (en) | 2013-05-06 | 2013-05-06 | Electronic telescope system based on image identification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013202417763U CN203299453U (en) | 2013-05-06 | 2013-05-06 | Electronic telescope system based on image identification |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN203299453U true CN203299453U (en) | 2013-11-20 |
Family
ID=49575453
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013202417763U Expired - Fee Related CN203299453U (en) | 2013-05-06 | 2013-05-06 | Electronic telescope system based on image identification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN203299453U (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103257442A (en) * | 2013-05-06 | 2013-08-21 | 深圳市中视典数字科技有限公司 | Electronic telescope system based on image identification and image processing method thereof |
-
2013
- 2013-05-06 CN CN2013202417763U patent/CN203299453U/en not_active Expired - Fee Related
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103257442A (en) * | 2013-05-06 | 2013-08-21 | 深圳市中视典数字科技有限公司 | Electronic telescope system based on image identification and image processing method thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109741369B (en) | Method and system for robot to track target pedestrian | |
| US20190034714A1 (en) | System and method for detecting hand gestures in a 3d space | |
| KR101628276B1 (en) | System and method for pathological analysis based on cloud | |
| CN112732071B (en) | Calibration-free eye movement tracking system and application | |
| CN110736748A (en) | Immunohistochemical nuclear plasma staining section diagnosis method and system | |
| CN110287907B (en) | Object detection method and device | |
| CN108090428B (en) | Face recognition method and system | |
| US20220012884A1 (en) | Image analysis system and analysis method | |
| CN111239999B (en) | Optical data processing method and device based on microscope and storage medium | |
| CN112507840B (en) | Man-machine hybrid enhanced small target detection and tracking method and system | |
| CN110533654A (en) | The method for detecting abnormality and device of components | |
| CN107609475B (en) | Pedestrian detection false detection extraction method based on light field camera | |
| US11361534B2 (en) | Method for glass detection in real scenes | |
| CN113470041B (en) | Immunohistochemical cell image cell nucleus segmentation and counting method and system | |
| CN112288699A (en) | Method, device, equipment and medium for evaluating relative definition of image | |
| CN110414558A (en) | Characteristic point matching method based on event camera | |
| CN203299453U (en) | Electronic telescope system based on image identification | |
| KR101509593B1 (en) | Image classification method and apparatus for preset tour camera | |
| CN103257442A (en) | Electronic telescope system based on image identification and image processing method thereof | |
| CN116486398A (en) | Focal image extraction method and device, electronic equipment and storage medium | |
| EP4270926A1 (en) | Image processing apparatus, image capture apparatus, image processing method, and program | |
| CN114004891A (en) | Distribution network line inspection method based on target tracking and related device | |
| CN113642425A (en) | Multi-mode-based image detection method and device, electronic equipment and storage medium | |
| CN106650252A (en) | Data processing method and device | |
| CN112991419A (en) | Parallax data generation method and device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20131120 Termination date: 20140506 |