CN105200938A - Vision-based anti-collision system for gate rail - Google Patents
Vision-based anti-collision system for gate rail Download PDFInfo
- Publication number
- CN105200938A CN105200938A CN201510534043.2A CN201510534043A CN105200938A CN 105200938 A CN105200938 A CN 105200938A CN 201510534043 A CN201510534043 A CN 201510534043A CN 105200938 A CN105200938 A CN 105200938A
- Authority
- CN
- China
- Prior art keywords
- image
- hash
- processing module
- average
- instant
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Image Processing (AREA)
Abstract
The invention relates to the field of gate rail control systems, and discloses a vision-based anti-collision system for a gate rail. The vision-based anti-collision system comprises a gate, a processing module, an image collecting module, a memory module and a host computer, wherein an image of a target region is collected on the basis of the image collecting module; the processing module is used for generating a hash fingerprint of the corresponding image through a hash algorithm, comparing a template image at the beginning of judgment with an instant image through the hash fingerprint, and obtaining whether a vehicle is in the instant image or not, so as to determine whether a gate rail is put down or not. The image is processed through the hash algorithm, so that the vision-based anti-collision system for the gate rail has the advantages of good robustness, good instantaneity, high stability, low algorithm complexity and small illumination variation effect; whether the vehicle is only in the target region of the lane or not can be effectively judged; the collision of the gate rail and the vehicle is effectively prevented.
Description
Technical field
The present invention relates to the track restrictor bar fender system in restrictor bar control system field, track, particularly a kind of view-based access control model.
Background technology
Freeway toll station is the important applied field of intelligent transportation system, in order to make the functional intellectualization of automobile and road, improves conevying efficiency, ensure traffic safety, alleviate congested in traffic, reduce environment pollution, prevent that the system of pounding seems particularly important in real time, accurately and efficiently.Because the chassis height of different vehicle is different, tenor is also different, the range of sensitivity of traditional ground induction coil is difficult to determine, do not fallen after causing vehicle also not passed through by just fall bar or vehicle the faults such as bar.In addition, ground induction coil can produce deformation under the spreading of oversize vehicle, even can cause not recoverable damage, thus causes the stability of induction coil to decline.
Summary of the invention
Goal of the invention of the present invention is: for above-mentioned technical problem, a kind of track restrictor bar fender system of view-based access control model is provided, make up the deficiency that ground induction coil exists, propose and pound system schema based on the anti-of image recognition, the stability of this programme is high, real-time good, antijamming capability is strong, has very large using value.
Technical solution of the present invention is: a kind of track restrictor bar fender system of view-based access control model, comprises banister, processing module, image capture module, memory module and host computer; Described image capture module is arranged on by banister, and just to target area; The image information that described processing module collects according to image capture module is analyzed, and controls the state of pedestrian guardrail; Store template image in described memory module, described processing module receives banister that track host computer sends when playing bar signal, and the image information that image capture module collects by control storage module is updated to template image; Processing module is carried out gaussian filtering to image information and is obtained instant image; Described processing module generates the Hash fingerprint of instant image and template image by hash algorithm, and Hash fingerprint is the iamge description character string combined with same order; When the Hash fingerprint Hamming distance of instant image and template image is greater than setting value, banister keeps railing to erect; When both Hamming distances are less than setting value, pedestrian guardrail is put down.
The present invention is based on the Treatment Analysis method of image, specifically image capture module gathers the image of target area, processing module generates the Hash fingerprint of correspondence image by hash algorithm, by to Hash fingerprint multilevel iudge template image at first and instant image, obtain whether having car in instant image, to determine whether pedestrian guardrail is put down.
Preferably, described processing module generates the Hash fingerprint of instant image and the Hash fingerprint of template image by DCT hash algorithm, the Hamming distance M of both calculating; Described processing module generates the Hash fingerprint of instant image and the Hash fingerprint of template image by average hash algorithm, the Hamming distance N of both calculating; When M and N is all greater than setting value, banister keeps valve rod to erect; When M and N is all less than setting value, banister valve rod puts down.
Here by two kinds of hash algorithms, image is analyzed simultaneously, effectively can avoid the incompleteness of algorithm, the antijamming capability of algorithm can be strengthened, and then the reliability of boosting algorithm, stability.
Preferably, described processing module generates 64 Hash fingerprints of instant image and 64 Hash fingerprints of template image by DCT hash algorithm, the Hamming distance M of both calculating; Described processing module generates 64 Hash fingerprints of instant image and 64 Hash fingerprints of template image by average hash algorithm, the Hamming distance N of both calculating; If M > 10 and N > 20, banister keeps railing to erect; If M < 5 and N < 10, pedestrian guardrail is put down.
Preferably, described hash algorithm is DCT hash algorithm, and the Hash fingerprint generation step of described instant image or template image is:
1) by the size of image down to 32 × 32,
2) image after reducing is converted into gray level image,
3) described gray level image is carried out dct transform, and obtains the DCT coefficient matrix of 32 × 32,
4) 8 × 8 matrixes in the described DCT coefficient matrix upper left corner are retained,
5) calculation procedure 4) Coefficient Mean in DCT coefficient matrix,
6) by step 4) in DCT matrix coefficient compare one by one with average, if be more than or equal to average, be designated as 1; If be less than average, be designated as 0; Comparative result is combined in order, forms 64 Hash fingerprints.
Whether DCT hash algorithm or make DCT perception hash algorithm can go out difference between two images by rapid identification here, specifically distinguishes difference between " having car " image, " without car " image, thus pick out in image and have car to exist.
Preferably, when the Hash fingerprint Hamming distance of instant image and template image is all greater than 10, banister keeps railing to erect; When both Hamming distances are less than equal 5, pedestrian guardrail is put down.
Preferably, described hash algorithm is average hash algorithm, and the Hash fingerprint generation step of described instant image or template image is:
1) by the size of image down to 8 × 8,64 pixels altogether;
2) convert the image of 8 × 8 to gray level image, and transfer 64 grades of gray scales to;
3) calculation procedure 2) average gray of all 64 pixels;
4) if will be more than or equal to average, be designated as 1; If be less than average, be designated as 0; Comparative result is combined in order, forms 64 Hash fingerprints.
Here average hash algorithm or be average perception hash algorithm, algorithm simple and fast, whether can go out the difference of two images under same background by rapid identification, be specifically distinguish difference between " having car " image, " without car " image, thus pick out in image and have car to exist.
Preferably, when the Hash fingerprint Hamming distance of instant image and template image is all greater than 20, banister keeps railing to erect; When both Hamming distances are all less than 10, pedestrian guardrail is put down.
Preferably, described image capture module comprises the ccd image sensor and SAA7113 video decoding chip that connect successively; Described processing module adopts DM642 chip; Described SAA7113 video decoding chip connects the VP1 interface of DM642 chip, and described DM642 chip passes through I
2c bus connection control SAA7113 video decoding chip.
Beneficial effect of the present invention is:
The present invention is processed image by hash algorithm, has that robustness is good, real-time is good, stability is high, algorithm complex is low and illumination variation affects little advantage; Effectively can judge that there is car a target area, track without car situation, effectively prevent restrictor bar to vehicle collision.
The present invention effectively invariably can respond to the defect existed in application by induction coil at road, provide a kind of and respond to efficient, accurate and durable track restrictor bar fender system.
Accompanying drawing explanation
Fig. 1 is high-level schematic functional block diagram of the present invention;
Fig. 2 is workflow schematic diagram of the present invention;
Fig. 3 is DCT coefficient matrix;
Fig. 4 is Hash fingerprint example 1;
Fig. 5 is Hash fingerprint example 2.
Wherein, 1-image capture module, 2-processing module, 3-memory module, 4-host computer, 5-banister, 6-power supply.
Detailed description of the invention
The invention discloses a kind of track restrictor bar fender system of view-based access control model, comprise banister, processing module, image capture module, memory module and host computer; Described image capture module is arranged on by banister, and just to target area; The image information that described processing module collects according to image capture module is analyzed, and controls the state of pedestrian guardrail; Store template image in described memory module, described processing module receives banister that track host computer sends when playing bar signal, and the image information that image capture module collects by control storage module is updated to template image; Processing module is carried out gaussian filtering to image information and is obtained instant image; Described processing module generates the Hash fingerprint of instant image and template image by hash algorithm, and Hash fingerprint is the iamge description character string combined with same order; When the Hash fingerprint Hamming distance of instant image and template image is greater than setting value, banister keeps railing to erect; When both Hamming distances are less than setting value, pedestrian guardrail is put down.
Here by hash algorithm, image is processed, by the difference between image, fast, effectively pick out in image and whether have car to exist, thus control pedestrian guardrail state, avoid pedestrian guardrail to pound bumper car.Specifically image capture module gathers the image of target area, processing module generates the Hash fingerprint of correspondence image by hash algorithm, by to Hash fingerprint multilevel iudge template image at first and instant image, obtain whether having car in instant image, to determine whether banister valve rod puts down.The more existing anti-system of pounding of the present invention decreases the application of induction coil, and structure is more stable and compact, and the efficiency of system in application, accuracy, stability all have significant raising than the existing anti-system of pounding.
Below in conjunction with accompanying drawing, the invention process is described.
As shown in Figure 1, be high-level schematic functional block diagram of the present invention, comprise image capture module 1, processing module 2, memory module 3, host computer 4, banister 5 and power supply 6.
Image capture module 1, for gathering the image information of target area, target area specifically direction of traffic is positioned at the region, track in railing front.Image capture module 1 comprises ccd image sensor and SAA7113 video decoding chip; The decoded data signal that CCD collects by SAA7113 becomes " VPO " data signal of standard and outputs to processing module 2, is equivalent to one " A/D " device.Wherein, processing module 2 also passes through I
2c twin wire universal serial bus connection control SAA7113 video decoding chip.
Processing module 2, for the analysis to picture signal, and connection control banister 5 railing state.Processing module 2 adopts DM642 (full name TMS320DM642) in TI company C6000 series DSP, DM642 core is C6416 type high-performance digital signal processor, there is extremely strong handling property, the flexibility of height and programmability, periphery is integrated with equipment and the interfaces such as very complete audio frequency, video and network service simultaneously, can meet to video/image process in native system, and complete relevant device connection control.
Memory module 3, specifically peripheral storage module, be connected with the EMIF interface of DM642 by external bus.Memory module 3 comprises SDRAM synchronous DRAM and FLASH solid-state memory.
Host computer 4 comprises functions such as opening local picture, the setting of IP address, reception image, select target region, sending zone coordinate, preservation image.
Banister 5, the entrance and exit of the passage management limiting motor-driven vehicle going on road is arranged.DM642 connection control banister 5 railing, the specifically motor of connection control railing, completes banister 5 railing state and controls.
Power supply 6, for system provides supply of electric power, specifically can use independently power-supply system.
As shown in Figure 2, be workflow schematic diagram of the present invention.Specific works step is as follows:
1. system initialization, comprises template image in memory module and initializes.
2. whether processing module monitoring host computer plays bar signal to sending banister.If send, carry out next step; If do not send, continue monitoring.
3. processing module receives the image information that image capture module sends, and this image information is updated to template image in memory module, and template image locks.
4. pair image information carries out gaussian filtering, obtains instant image.
5. processing module carries out Treatment Analysis to instant image, specifically based on hash algorithm process, comprises DCT (perception) hash algorithm and average (perception) hash algorithm.
Wherein, DCT hash algorithm flow process comprises:
1) minification: by instant image down to the size of 32 × 32, simplify the calculating of DCT.
2) color is simplified: the instant image reduced is changed into gray level image, simplifies amount of calculation further.
3) DCT is calculated: the dct transform calculating picture: F (u, v)=C
tfC, obtains the DCT coefficient matrix of 32 × 32:
4) reduce DCT: according to 3) dct transform formula F (u, v)=C
tfC can obtain the DCT coefficient matrix that size is 32 × 32, and retains 8 × 8 matrixes in the DCT coefficient matrix upper left corner, and this part presents the low-limit frequency in picture; As shown in Figure 3.
5) calculating mean value: the average calculating DCT, as:
6) Hash fingerprint is generated: the DCT coefficient of 8 × 8 and average are compared, be more than or equal to average, be designated as 1, be less than average, be designated as 0; As shown in Figure 4.Finally comparative result is combined in order, just constitute the Hash fingerprint of 64.Here it should be noted that, the Hash fingerprint of instant image and template image needs identical sequential arrangement, as Fig. 4 according to from top to bottom often row be arranged in order fingerprint be: 11,100,000,100,110,001,000,001,000,100,000,100,000,000,000,000,010,001,000 00100000.
Average hash algorithm flow process comprises:
1) minification: by the size of instant image down to 8 × 8,64 pixels altogether.
2) color is simplified: convert the instant image of 8 × 8 to gray level image, and transfer 64 grades of gray scales to.
3) calculating mean value: the average gray calculating all 64 pixels; Such as:
4) gray scale of compared pixels: the gray scale of each pixel and average are compared, is more than or equal to average, be designated as 1, be less than average, be designated as 0; As shown in Figure 5.
5) generate Hash fingerprint: by 4) comparative result, combine according to a certain order, just constitute the Hash fingerprint of 64.As Fig. 5 according to from top to bottom often row be arranged in order fingerprint be: 11,111,111,111,111,111,101,111,111,011,011,110,000,111,100,001,111,000,011 11100011.
6. the Hash fingerprint that the Hash fingerprint obtained by instant image and template image obtain contrasts, when the Hamming distance of DCT perception Hash fingerprint is greater than 10, and the Hamming distance of average perception Hash fingerprint is when being greater than 20, illustrate in video area to there is vehicle, now control restrictor bar and keep vertical state.
7. processing module receives the image information repetition step 4-6 that image capture module sends, until the Hamming distance of DCT perception Hash fingerprint is less than 5, and when the Hamming distance of average perception Hash fingerprint is less than 10, just controls restrictor bar and falls.
The present invention can reduce the application of ground induction coil on system composition, can avoid the defect that it exists in road induction like this; Meanwhile, by two kinds of hash algorithms, image is analyzed simultaneously, the reasonability of algorithm can be ensured, the antijamming capability of algorithm can be strengthened, and then the reliability of boosting algorithm and system, stability.
Claims (8)
1. a track restrictor bar fender system for view-based access control model, comprises banister, processing module, image capture module, memory module and host computer; Described image capture module is arranged on by banister, and just to target area; The image information that described processing module collects according to image capture module is analyzed, and controls the state of pedestrian guardrail; It is characterized in that: store template image in described memory module, described processing module receives banister that track host computer sends when playing bar signal, and the image information that image capture module collects by control storage module is updated to template image; Processing module is carried out gaussian filtering to image information and is obtained instant image; Described processing module generates the Hash fingerprint of instant image and template image respectively by hash algorithm, and Hash fingerprint is the iamge description character string combined with same order; When the Hash fingerprint Hamming distance of instant image and template image is greater than setting value, banister keeps railing to erect; When both Hamming distances are less than setting value, pedestrian guardrail is put down.
2. the track restrictor bar fender system of a kind of view-based access control model according to claim 1, is characterized in that: described processing module generates the Hash fingerprint of instant image and the Hash fingerprint of template image by DCT hash algorithm, compares both line Hamming distance M; Described processing module generates the Hash fingerprint of instant image and the Hash fingerprint of template image by average hash algorithm, compares both line Hamming distance N; When M and N is all greater than setting value, banister keeps railing to erect; When M and N is all less than setting value, pedestrian guardrail is put down.
3. the track restrictor bar fender system of a kind of view-based access control model according to claim 2, it is characterized in that: described processing module generates 64 Hash fingerprints of instant image and 64 Hash fingerprints of template image by DCT hash algorithm, the Hamming distance M of both calculating; Described processing module generates 64 Hash fingerprints of instant image and 64 Hash fingerprints of template image by average hash algorithm, the Hamming distance N of both calculating; If M > 10 and N > 20, banister keeps railing to erect; If M < 5 and N < 10, pedestrian guardrail is put down.
4. the track restrictor bar fender system of a kind of view-based access control model according to claim 1, is characterized in that: described hash algorithm is DCT hash algorithm, and the Hash fingerprint generation step of described instant image or template image is:
1) by the size of image down to 32 × 32,
2) image after reducing is converted into gray level image,
3) described gray level image is carried out dct transform, and obtains the DCT coefficient matrix of 32 × 32,
4) 8 × 8 matrixes in the described DCT coefficient matrix upper left corner are retained,
5) calculation procedure 4) Coefficient Mean in DCT coefficient matrix,
6) by step 4) in DCT matrix coefficient compare one by one with average, if be more than or equal to average, be designated as 1; If be less than average, be designated as 0; Comparative result is combined in order, forms 64 Hash fingerprints.
5. the track restrictor bar fender system of a kind of view-based access control model according to claim 4, is characterized in that: if the Hash fingerprint Hamming distance of instant image and template image is all greater than 10, then banister keeps railing to erect; If when both Hamming distances are all less than 5, then pedestrian guardrail is put down.
6. the track restrictor bar fender system of a kind of view-based access control model according to claim 1, is characterized in that: described hash algorithm is average hash algorithm, and the Hash fingerprint generation step of described instant image or template image is:
1) by the size of image down to 8 × 8,64 pixels altogether;
2) convert the image of 8 × 8 to gray level image, and transfer 64 grades of gray scales to;
3) calculation procedure 2) average gray of all 64 pixels;
4) by step 2) in 64 grades of gray scales and average gray compare one by one, if be more than or equal to average, be designated as 1; If be less than average, be designated as 0; Comparative result is combined in order, forms 64 Hash fingerprints.
7. the track restrictor bar fender system of a kind of view-based access control model according to claim 6, is characterized in that: if when the Hash fingerprint Hamming distance of instant image and template image is all greater than 20, then banister keeps railing to erect; If when both Hamming distances are all less than 10, then pedestrian guardrail is put down.
8. the track restrictor bar fender system of a kind of view-based access control model according to claim 1, is characterized in that: described image capture module comprises the ccd image sensor and SAA7113 video decoding chip that connect successively; Described processing module adopts DM642 chip; Described SAA7113 video decoding chip connects the VP1 interface of DM642 chip, and described DM642 chip passes through I
2c bus connection control SAA7113 video decoding chip.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510534043.2A CN105200938B (en) | 2015-08-27 | 2015-08-27 | A kind of track restrictor bar CAS of view-based access control model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510534043.2A CN105200938B (en) | 2015-08-27 | 2015-08-27 | A kind of track restrictor bar CAS of view-based access control model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105200938A true CN105200938A (en) | 2015-12-30 |
| CN105200938B CN105200938B (en) | 2017-06-27 |
Family
ID=54948871
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510534043.2A Active CN105200938B (en) | 2015-08-27 | 2015-08-27 | A kind of track restrictor bar CAS of view-based access control model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105200938B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106403951A (en) * | 2016-08-26 | 2017-02-15 | 观宇能源科技（上海）有限公司 | Computer vision based positioning system and positioning method thereof |
| CN109636797A (en) * | 2018-12-21 | 2019-04-16 | 国网福建省电力有限公司 | A kind of method of contactless flexible direct current converter valve module shape changing detection |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103336890A (en) * | 2013-06-08 | 2013-10-02 | 东南大学 | Method for quickly computing similarity of software |
| CN103343515A (en) * | 2013-07-16 | 2013-10-09 | 东北石油大学 | Anti-collision intelligent road brake and intelligent recognition control method thereof |
| CN103780899A (en) * | 2012-10-25 | 2014-05-07 | 华为技术有限公司 | Method and device for detecting whether camera is interfered and video monitoring system |
| US20140223565A1 (en) * | 2012-08-29 | 2014-08-07 | The Johns Hopkins University | Apparatus And Method For Identifying Similarity Via Dynamic Decimation Of Token Sequence N-Grams |
| CN104866616A (en) * | 2015-06-07 | 2015-08-26 | 中科院成都信息技术股份有限公司 | Method for searching monitor video target |
-
2015
- 2015-08-27 CN CN201510534043.2A patent/CN105200938B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140223565A1 (en) * | 2012-08-29 | 2014-08-07 | The Johns Hopkins University | Apparatus And Method For Identifying Similarity Via Dynamic Decimation Of Token Sequence N-Grams |
| CN103780899A (en) * | 2012-10-25 | 2014-05-07 | 华为技术有限公司 | Method and device for detecting whether camera is interfered and video monitoring system |
| CN103336890A (en) * | 2013-06-08 | 2013-10-02 | 东南大学 | Method for quickly computing similarity of software |
| CN103343515A (en) * | 2013-07-16 | 2013-10-09 | 东北石油大学 | Anti-collision intelligent road brake and intelligent recognition control method thereof |
| CN104866616A (en) * | 2015-06-07 | 2015-08-26 | 中科院成都信息技术股份有限公司 | Method for searching monitor video target |
Non-Patent Citations (1)
| Title |
|---|
| 王永等: "基于双目视觉的车辆闸杆防撞***", 《计算机***应用》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106403951A (en) * | 2016-08-26 | 2017-02-15 | 观宇能源科技（上海）有限公司 | Computer vision based positioning system and positioning method thereof |
| CN109636797A (en) * | 2018-12-21 | 2019-04-16 | 国网福建省电力有限公司 | A kind of method of contactless flexible direct current converter valve module shape changing detection |
| CN109636797B (en) * | 2018-12-21 | 2022-08-05 | 国网福建省电力有限公司 | Method for detecting deformation of non-contact flexible direct current converter valve module |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105200938B (en) | 2017-06-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109460709B (en) | RTG visual barrier detection method based on RGB and D information fusion | |
| CN103500322B (en) | Automatic lane line identification method based on low latitude Aerial Images | |
| CN106951879B (en) | Multi-feature fusion vehicle detection method based on camera and millimeter wave radar | |
| CN104036262B (en) | A kind of method and system of LPR car plates screening identification | |
| CN110298300B (en) | Method for detecting vehicle illegal line pressing | |
| CN101980245B (en) | Adaptive template matching-based passenger flow statistical method | |
| CN105225281B (en) | A kind of vehicle checking method | |
| CN104463235A (en) | Fault recognition method and device based on operation images of motor train unit | |
| CN109489724A (en) | A kind of tunnel safe train operation environment comprehensive detection device and detection method | |
| CN107993488A (en) | A kind of parking stall recognition methods, system and medium based on fisheye camera | |
| CN111311766A (en) | Roadside parking intelligent charging system and method based on license plate recognition and tracking technology | |
| CN112818834B (en) | Method, device and medium for judging avoidance of emergency vehicle at intersection | |
| CN112084928A (en) | Road traffic accident detection method based on visual attention mechanism and ConvLSTM network | |
| CN103337175A (en) | Vehicle type recognition system based on real-time video steam | |
| CN114494161A (en) | Pantograph foreign matter detection method and device based on image contrast and storage medium | |
| CN114255428A (en) | Vehicle parking identification method based on edge intelligence and roadside high-level monitoring video | |
| CN105200938A (en) | Vision-based anti-collision system for gate rail | |
| CN108921147B (en) | Black smoke vehicle identification method based on dynamic texture and transform domain space-time characteristics | |
| CN106960193A (en) | A kind of lane detection apparatus and method | |
| CN114092857A (en) | Gateway-based collection card image acquisition method, system, equipment and storage medium | |
| CN112053407B (en) | Automatic lane line detection method based on AI technology in traffic law enforcement image | |
| Shafie et al. | Smart video surveillance system for vehicle detection and traffic flow control | |
| CN105631425A (en) | License plate recognition method based on video streams and system thereof and intelligent digital video camera | |
| CN117058912A (en) | Method and device for detecting abnormal parking of inspection vehicle, storage medium and electronic equipment | |
| Tan et al. | Shape template based side-view car detection algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |