CN118038165A - Video-based automatic identification algorithm for black smoke event of ship - Google Patents
Video-based automatic identification algorithm for black smoke event of ship Download PDFInfo
- Publication number
- CN118038165A CN118038165A CN202410228256.1A CN202410228256A CN118038165A CN 118038165 A CN118038165 A CN 118038165A CN 202410228256 A CN202410228256 A CN 202410228256A CN 118038165 A CN118038165 A CN 118038165A
- Authority
- CN
- China
- Prior art keywords
- ship
- black smoke
- video
- identification
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000000779 smoke Substances 0.000 title claims abstract description 147
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 7
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000012216 screening Methods 0.000 claims description 4
- 230000007547 defect Effects 0.000 abstract description 6
- 238000000034 method Methods 0.000 description 19
- 238000012545 processing Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses an automatic identification algorithm of a ship black smoke event based on video, which comprises the following steps: s1, video acquisition is carried out on a designated water area, black smoke identification and ship identification are respectively carried out on each frame of image in the video, and a marked image set of a black smoke marked black smoke identification frame and a ship marked ship identification frame in the image is obtained; s2, identifying whether a ship black smoke event exists in the video; s3, based on the starting time and the ending time of black smoke emission of the ship i with the primary illegal ship black smoke emission event judged in the step S2, extracting illegal videos of the ship i from the video acquired in the step S1; the automatic recognition algorithm is not affected by unstable exhaust emission of the ship and defects of the black smoke recognition algorithm, can effectively remove part of wrong black smoke recognition results, and has the advantages of simplicity, high efficiency and high accuracy.
Description
Technical Field
The invention relates to the technical field of ship tail gas monitoring, in particular to an automatic identification algorithm for a ship black smoke event based on video.
Background
The black smoke of the ship is a serious to macroscopic ship atmospheric pollution phenomenon, and at present, marine law enforcement personnel mainly judge the blackness of the tail gas of the ship by means of naked eye observation and contrast with a ringeman blackness map, so that the problem of low efficiency exists. With the development of artificial intelligence technology and the popularization of video monitoring systems, the use of artificial intelligence to identify ships or black smoke in pictures is technically feasible; however, from the viewpoint of marine supervision, the transient black smoke phenomenon caused by temporary acceleration or deceleration of the ship does not belong to the illegal behavior in actual operation, and only the continuous black smoke phenomenon of the ship in a constant-speed sailing state belongs to the illegal behavior. Therefore, the ship black smoke snapshot system not only needs to identify each frame of image in the video and accurately judge whether black smoke exists or not, or other similar phenomena, but also needs to judge whether a multi-frame ship black smoke image belongs to a single ship black smoke event or two or more independent black smoke events from different ships.
Because the ship tail gas belongs to a non-rigid target, the method has the characteristics of unstable shape and varying blackness, and compared with the identification of the ship in the image, the method has the advantages that the error identification rate and the missing identification rate of black smoke emitted by the ship in the image are large and difficult to avoid, and a great challenge is brought to the automatic identification of the subsequent black smoke emission event of the ship in the current image identification field. Therefore, how to accurately identify the illegal black smoke emission event of the ship can still be realized under the condition of more misidentification rate and missed identification of the tail gas of the ship is a key for developing a ship black smoke snapshot and identification system.
Disclosure of Invention
The invention aims to provide an automatic identification algorithm for black smoke emission events of ships based on videos, which solves the technical problems.
For this purpose, the technical scheme of the invention is as follows:
the automatic identification algorithm for the black smoke event of the ship based on the video comprises the following steps:
S1, video acquisition is carried out on a designated water area, black smoke identification and ship identification are respectively carried out on each frame of image in the video, and a marked image set of a black smoke marked black smoke identification frame and a ship marked ship identification frame in the image is obtained;
s2, identifying whether a ship black smoke event exists in the video, wherein the specific steps are as follows:
s201, respectively calculating the center distances d i,t between each black smoke identification frame and different ship identification frames in each marked image, wherein the calculation formula is as follows:
di,t＝psj,t-pvi,t，
Wherein t is the acquisition time of the marked image; ps j,t is the pixel abscissa of the center point of the black smoke identification frame; j is the serial number of the black smoke identification frame in the marked image, and pv i,t is the pixel abscissa of the center point of the ship identification frame; i is the number of the ship, different ships have different numbers, and the same ship has the same number in different frames of images;
S202, calculating an average value D i of all center point distances D i,t of the same ship i;
S203, calculating a distance deviation absolute value b i,t corresponding to the distance d i,t between the center points of the same ship i, wherein the calculation formula is as follows:
bi,t＝∣di,t-Di∣；
S204, counting the times n i when the absolute value b i,t of the distance deviation in the same ship i is smaller than the maximum distance deviation upper limit value b max according to the set maximum distance deviation upper limit value b max;
S205, according to the set minimum number of times of identifying ship black smoke, n min, and comparing the minimum number of times of identifying ship black smoke with n i: when n i≥nmin, the ship i has a black smoke emission event of a illegal ship; when n i＜nmin is reached, the ship i has no illegal ship black smoke emission event;
And S3, screening out absolute values b i,t of a plurality of distance deviations which are smaller than a maximum distance deviation upper limit value b max from absolute values b i,t of all distance deviations corresponding to a ship i with one-time offending ship black smoke event judged in the step S2, and finding out a minimum value t min and a maximum value t max of a marked image acquisition moment t in the absolute values b i,t, so that offending videos of the ship i are extracted in the step S1 by taking t min as video start time and t max as video end time.
Further, in step S1, the image acquisition frequency of the video is 2 frames/second to 5 frames/second.
Further, in step S1, black smoke identification and marking in the image are performed by a black smoke identification algorithm, and ship identification and marking in the image are performed by a ship identification algorithm.
Further, in step 4) of step S2, the maximum distance deviation upper limit value b max is set to 50 to 100.
Further, in step 5) of step S2, the minimum number of times n min of identifying the ship black smoke is set to 10 to 100 times.
According to the automatic identification algorithm for the black smoke event of the ship based on the video, under the condition that the image identification error identification rate and the missing identification rate of the tail gas of the current ship are high, the identification defect of the black smoke of the tail gas of the ship is compensated by utilizing the characteristics that the ship belongs to a rigid target and the tail gas of the ship moves relative to the ship, so that the accurate identification of the black smoke of the tail gas of the ship is realized, and the illegal black smoke event judgment of the ship is completed based on the identification defect.
Compared with the prior art, the video-based ship black smoke event automatic identification algorithm has the beneficial effects that: 1) The method is not influenced by unstable exhaust emission of the ship, and the same ship can emit black smoke indirectly for many times and can be accurately identified as a complete ship black smoke emission event; 2) The method has the advantages that the identification of the ship black smoke event is not affected by the defect of a black smoke identification algorithm, and the final identification result is not affected even if the black smoke with partial form and color in the video is not accurately identified; 3) The method can effectively remove partial wrong black smoke identification results (such as cloud bottom shadows in cloudy days, smoke discharged by remote fixed pollution sources, some evening canberat in evening, projection of high buildings on the ship body at the river side, and the like); in conclusion, the method has the advantages of simplicity, high efficiency and high accuracy, and has good application and popularization prospects for marine supervision departments in realizing automatic identification of black smoke events of ships based on videos.
Drawings
FIG. 1 is a flow chart of the video-based automatic identification algorithm for black smoke event of a ship of the present invention;
FIG. 2 is a schematic diagram of a marked image obtained in step S1 of the video-based automatic identification algorithm for black smoke event of a ship according to the present invention;
Fig. 3 is a schematic diagram of a marked image of a mismarked building projected as black smoke on a ship body based on the existing black smoke recognition algorithm in step S1 of the automatic recognition algorithm of the black smoke event of the ship based on video of the invention;
Fig. 4 is a schematic diagram of a marked image with cloud-background shadows as black smoke, which is obtained by mismarking based on the existing black smoke recognition algorithm in step S1 of the automatic recognition algorithm for the black smoke event of the ship based on the video.
Detailed Description
The invention will now be further described with reference to the accompanying drawings and specific examples, which are in no way limiting.
Example 1
Referring to fig. 1, the embodiment of the present application is implemented based on an image acquired by a camera installed at a certain wharf in a sea area of the Yangtze river at 1 month 9 of 2024, and the implementation steps are as follows:
s1, a camera arranged at the high position of a wharf collects video of a designated water area at an image collection frequency of 2 frames/second, and respectively performs black smoke identification and ship identification on each frame of image in the video to obtain a marked image set of a black smoke marked black smoke identification frame and a ship marked ship identification frame in the image, and the marked image set is shown in fig. 2.
The ship identification and marking in the image are completed through the existing black smoke identification algorithm, specifically, the ship in the image is identified, and the ship is marked by a minimum circumscribed rectangular frame, namely a ship identification frame; the black smoke identification and marking in the image are completed through the existing black smoke identification algorithm, specifically, the black smoke part in the image is identified, and the black smoke part is marked by a minimum circumscribed rectangular frame, namely the black smoke identification frame; because the dark cloud, the black reflection of an object and the like in the type of image are similar to the image of the black smoke of the tail gas of the ship, the problem of false identification is easy to occur; when the color difference between the black smoke color and the natural color at the time of image acquisition is smaller, the problem of missing identification of the black smoke of the tail gas of the ship is easy to occur;
In practical application, black smoke recognition and ship recognition are respectively and independently carried out based on two image recognition algorithms, and marked in images respectively; because the images are acquired by the same camera and have the same resolution, a pixel coordinate system is constructed for the images based on the images, and the column number and the row number of the pixel units in the images are utilized to represent the coordinates of the points in the images, so that the center point abscissa ps j,t of the black smoke identification frame and the center point abscissa pv i,t of the ship identification frame in each image are obtained;
s2, identifying whether a ship black smoke event exists in the video, wherein the specific steps are as follows:
In the video acquired by the camera at 2024 1 month 9, automatically identifying 39 frames of images of the same ship from 7 hours, 6 minutes, 16 seconds and 36.5 seconds; wherein, because of the error recognition and missing recognition problems of the existing black smoke recognition algorithm, 39 times of ship recognition frames are recognized in the 39 frames of images, but only 31 times of black smoke recognition frames are recognized.
The center point pixel coordinates of the black smoke recognition frame and the ship recognition frame recognized at each acquisition time of the 39 frames of images are shown in table 1 below. In table 1, ps j,t columns of data are blank, that is, black smoke may not be recognized in the image acquired at the moment due to the missing recognition problem, and thus the center point coordinate data of the black smoke recognition frame is not available.
Table 1:
S201, according to the formula: di, t= psj, t-pvi, t, respectively calculating the center distances d i,t between each black smoke identification frame and different ship identification frames in each marked image, wherein t is the acquisition time of the marked image; ps j,t is the pixel abscissa of the center point of the black smoke identification frame; j is the serial number of the black smoke identification frame in the marked image, and pv i,t is the pixel abscissa of the center point of the ship identification frame; i is the ship number;
s202, calculating an average value D i of all center point distances D i,t of the ship i;
s203, according to the formula: b i,t＝∣di,t-Di | calculating a distance deviation absolute value b i,t corresponding to the distance d i,t between the central points of the ship i;
In this embodiment, the specific calculation result of the center distance d i,t between each black smoke identification frame and different ship identification frames in step S201 is shown in table 1; calculating to obtain an average value D i = -45.4 according to the center distance D i,t obtained in the step S201; referring to table 1 for the specific calculation result of the absolute value b i,t of the distance deviation corresponding to the center point distance d i,t in step S202, it can be seen from table 1 that the minimum value of the absolute value b i,t of the distance deviation is 0.4 and the maximum value thereof is 40.6;
S204, setting a maximum distance deviation upper limit value b max as 50, and counting the times n i that the absolute value b i,t of the distance deviation in the ship i is smaller than the maximum distance deviation upper limit value b max as 30 times;
s205, setting the minimum number of times of identifying black smoke of the ship as n min as 10 times, and judging that a black smoke event of the ship with one violation exists in the ship i due to n i＞nmin;
S3, screening out absolute values b i,t of a plurality of distance deviations which are smaller than a maximum distance deviation upper limit value b max from absolute values b i,t of all distance deviations corresponding to the ship i, finding out the minimum value t min and the maximum value t max of the marked image acquisition time t which are respectively 17 th second and 36.5 th second, and further storing videos from 17 th second to 36.5 th second in the step S1 acquisition video as illegal videos of the ship i.
In order to further prove the effectiveness of the application, a video record of 0 minutes at 7 to 10 minutes at 7 hours is called in the video acquired in the step S1, the running-in and running-out process of the ship i from 0 minutes 16 seconds to 36.5 seconds in the video can be clearly seen according to the video record, the scene of black smoke tail gas emitted by the ship can be clearly seen in the running-in and running-out process of the ship i, and the result accords with the result obtained by automatic identification in the embodiment; in addition, in the data processing process of the embodiment, although the existence of black smoke is not recognized in the 16 th second, 16.5 th second, 18 th second, 24.5 th second, 25 th second, 25.5 th second, 28 th second and 28.5 th second images, the fact that the ship i has a illegal ship black smoke event is accurately judged by the method is not influenced; in addition, referring to fig. 3, in the image processing of the video, the condition that the black smoke mark frame is wrongly marked on the projection of the building on the ship body exists, but the accuracy of the application data result is not affected, so that the method provided by the application is proved to be capable of accurately realizing the automatic identification of the illegal black smoke behavior of the ship under the condition that the black smoke identification of the tail gas of the ship is defective.
Example 2
Referring to fig. 1, the embodiment of the present application is implemented based on an image acquired by a camera installed at a wharf in a sea area on the Yangtze river at 22 days 1 month 2024, and the implementation steps are as follows:
S1, carrying out video acquisition on a designated water area in the same image acquisition mode as the embodiment, and respectively carrying out black smoke identification and ship identification on each frame of image in the video to obtain a marked image set of a black smoke marked black smoke identification frame and a ship marked ship identification frame in the image;
s2, identifying whether a ship black smoke event exists in the video, wherein the specific steps are as follows:
automatically recognizing 28 frames of images of the same ship from 16 hours, 9 minutes, 10 seconds and 25 seconds in video acquired by a camera at 22 days of 1 month of 2024; similarly, due to the frame loss problem, 28 times of ship identification frames are identified in the 28-frame image, but only 22 times of black smoke identification frames are identified.
The center point pixel coordinates of the black smoke recognition frame and the ship recognition frame recognized by the 28 frames of images at each acquisition time are shown in the following table 2, and the center point pixel coordinates are shown in the following table 2; similarly, there was a blank ps j,t column data, i.e., no black smoke was identified in the image acquired at that time.
Table 2:
S201, respectively calculating the center distances d i,t between each black smoke identification frame and different ship identification frames in each marked image, wherein the specific calculation results are shown in Table 2;
S202, calculating an average value D i,Di = -277.9 of all the center point distances D i,t of the ship i;
S203, calculating a distance deviation absolute value b i,t corresponding to the distance d i,t between the center points of the ship i, wherein the specific calculation result is shown in table 2; as can be seen from table 2, the minimum value of the absolute value b i,t of the distance deviation is 1.9, and the maximum value is 42.1;
S204, setting a maximum distance deviation upper limit value b max as 50, and counting the times n i that the absolute value b i,t of the distance deviation in the ship i is smaller than the maximum distance deviation upper limit value b max as 22 times;
s205, setting the minimum number of times of identifying black smoke of the ship as n min as 10 times, and judging that a black smoke event of the ship with one violation exists in the ship i due to n i＞nmin;
S3, screening out absolute values b i,t of a plurality of distance deviations which are smaller than a maximum distance deviation upper limit value b max from absolute values b i,t of all distance deviations corresponding to the ship i, finding out that the minimum value t min and the maximum value t max of the marked image acquisition time t are respectively 10 th second and 25 th second, and further storing the video from 10 th second to 25 th second in the acquired video in the step S1 as illegal video of the ship i.
In order to further prove the effectiveness of the application, a video record of 5 minutes to 15 minutes at 16 hours is called in the video acquired in the step S1, the process of entering and exiting the ship i in the video from 9 minutes, 10 seconds to 9 minutes, 25 seconds can be seen according to the video record, and the scene that the ship i emits obvious black smoke tail gas along with the 10 th to 25 th seconds can be cleared in the video, so that the method accords with the result obtained by automatic identification in the embodiment; it can be seen that, in the data processing process, although black smoke is not recognized in the 14.5 th to 17 th second images, the fact that the ship i has a illegal ship black smoke event is accurately judged by the method of the application is not affected; in addition, as can be seen from the minimum value and the maximum value in the distance deviation b i,t data obtained by the processing in step S203, although the initial data has the problem of larger deviation difference, the effective data can be screened by using the method; furthermore, the method provided by the application can accurately realize the automatic identification of the illegal black smoke emission behavior of the ship under the condition that the black smoke identification of the tail gas of the ship has defects.
Example 3
Referring to fig. 1, the embodiment of the present application is implemented based on an image acquired by a camera installed at a dock in a water area of a river in the yellow pu of the Shanghai at 22 days 1 month in 2024, and the implementation steps are as follows:
S1, carrying out video acquisition on a designated water area in the same image acquisition mode as the embodiment, and respectively carrying out black smoke identification and ship identification on each frame of image in the video to obtain a marked image set of a black smoke marked black smoke identification frame and a ship marked ship identification frame in the image;
s2, identifying whether a ship black smoke event exists in the video, wherein the specific steps are as follows:
Automatically recognizing 27 frames of images of the same ship from 25 minutes and 0 seconds to 14.5 seconds in the video acquired by the camera at 22 days 1 and 22 of 2024; similarly, due to the frame loss problem, 27 times of ship identification frames are identified in the 27 frames of images, but only 11 times of black smoke identification frames are identified.
The center point pixel coordinates of the black smoke recognition frame and the ship recognition frame recognized by the 27 frames of images at each acquisition time are shown in the following table 3, and the following table 3; similarly, there was a blank ps j,t column data, i.e., no black smoke was identified in the image acquired at that time.
Table 3:
S201, respectively calculating the center distances d i,t between each black smoke identification frame and different ship identification frames in each marked image, wherein the specific calculation results are shown in Table 3;
S202, calculating an average value D i,Di = -544.9 of all the center point distances D i,t of the ship i;
S203, calculating a distance deviation absolute value b i,t corresponding to the distance d i,t between the center points of the ship i, wherein the specific calculation result is shown in a table 3; as can be seen from table 3, the minimum value of the absolute value b i,t of the distance deviation is 110.1, and the maximum value is 745.1;
S204, setting a maximum distance deviation upper limit value b max as 50, and counting the times n i that the absolute value b i,t of the distance deviation in the ship i is smaller than the maximum distance deviation upper limit value b max as 0 times;
S205, setting the minimum number of times of identifying the black smoke of the ship n min as 10 times, and judging that the ship i has no illegal black smoke emission event due to n i＜nmin.
S3, as the ship i does not have a black smoke emission event of the illegal ship, the video without the violation needs to be stored.
In order to further prove the effectiveness of the application, the video records of 15 minutes to 35 minutes at 14 hours are called in the video collected in the step S1, the entering and exiting processes of the ship i can be seen according to the video records, and as shown in fig. 4, the cloud is always mistakenly identified as black smoke when the black smoke identification is carried out on the image in the embodiment, so that the pv i,t in the data of the table 3 is continuously enlarged from 119 to 1836, the position of a black smoke identification frame is basically unchanged, namely, p st fluctuates between 115 and 133, and therefore, the ship i does not have the problem of illegal black smoke emission of the ship; the method provided by the application can accurately remove the interference data under the condition that the black smoke identification of the tail gas of the ship has defects, so that the automatic identification of the illegal black smoke emission behavior of the ship is realized.
Claims (5)
1. The automatic identification algorithm for the black smoke event of the ship based on the video is characterized by comprising the following steps:
S1, video acquisition is carried out on a designated water area, black smoke identification and ship identification are respectively carried out on each frame of image in the video, and a marked image set of a black smoke marked black smoke identification frame and a ship marked ship identification frame in the image is obtained;
s2, identifying whether a ship black smoke event exists in the video, wherein the specific steps are as follows:
s201, respectively calculating the center distances d i,t between each black smoke identification frame and different ship identification frames in each marked image, wherein the calculation formula is as follows:
di,t＝psj,t-pvi,t，
wherein t is the acquisition time of the marked image; ps j,t is the pixel abscissa of the center point of the black smoke identification frame; j is the serial number of the black smoke identification frame in the marked image, and pv i,t is the pixel abscissa of the center point of the ship identification frame; i is the ship number;
S202, calculating an average value D i of all center point distances D i,t of the same ship i;
S203, calculating a distance deviation absolute value b i,t corresponding to the distance d i,t between the center points of the same ship i, wherein the calculation formula is as follows:
bi,t＝∣di,t-Di∣；
S204, counting the times n i when the absolute value b i,t of the distance deviation in the same ship i is smaller than the maximum distance deviation upper limit value b max according to the set maximum distance deviation upper limit value b max;
S205, according to the set minimum number of times of identifying ship black smoke, n min, and comparing the minimum number of times of identifying ship black smoke with n i: when n i≥nmin, the ship i has a black smoke emission event of a illegal ship; when n i＜nmin is reached, the ship i has no illegal ship black smoke emission event;
And S3, screening out absolute values b i,t of a plurality of distance deviations which are smaller than a maximum distance deviation upper limit value b max from absolute values b i,t of all distance deviations corresponding to a ship i with one-time offending ship black smoke event judged in the step S2, and finding out a minimum value t min and a maximum value t max of a marked image acquisition moment t in the absolute values b i,t, so that offending videos of the ship i are extracted in the step S1 by taking t min as video start time and t max as video end time.
2. The automatic identification algorithm for black smoke event of a video-based ship according to claim 1, wherein in step S1, the image acquisition frequency of the video is 2 frames/sec to 5 frames/sec.
3. The automatic recognition algorithm for the black smoke event of the ship based on the video according to claim 1, wherein in step S1, the black smoke recognition and the marking of the image are performed by the black smoke recognition algorithm, and the ship recognition and the marking of the image are performed by the ship recognition algorithm.
4. The automatic recognition algorithm of a black smoke event of a video-based ship according to claim 1, wherein in step 4) of step S2, the maximum distance deviation upper limit value b max is set to 50 to 100.
5. The automatic recognition algorithm of black smoke event of a ship based on video according to claim 1, wherein in step 5) of step S2, the minimum number of times n min of recognizing black smoke of a ship is set to 10 to 100 times.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410228256.1A CN118038165A (en) | 2024-02-29 | 2024-02-29 | Video-based automatic identification algorithm for black smoke event of ship |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410228256.1A CN118038165A (en) | 2024-02-29 | 2024-02-29 | Video-based automatic identification algorithm for black smoke event of ship |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118038165A true CN118038165A (en) | 2024-05-14 |
Family
ID=90993014
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410228256.1A Pending CN118038165A (en) | 2024-02-29 | 2024-02-29 | Video-based automatic identification algorithm for black smoke event of ship |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118038165A (en) |
-
2024
- 2024-02-29 CN CN202410228256.1A patent/CN118038165A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109190636B (en) | Remote sensing image ship target information extraction method | |
| CN110070530B (en) | Transmission line icing detection method based on deep neural network | |
| CN101553852B (en) | System and method for detection of average speed of vehicles for traffic control | |
| CN108806334A (en) | A kind of intelligent ship personal identification method based on image | |
| CN108334881B (en) | License plate recognition method based on deep learning | |
| CN110210428B (en) | MSER-based smoke root node detection method in remote complex environment | |
| CN109886978B (en) | End-to-end alarm information identification method based on deep learning | |
| CN110866926B (en) | Infrared remote sensing image rapid and fine sea-land segmentation method | |
| CN113139572B (en) | Image-based train air spring fault detection method | |
| CN113808282A (en) | Multi-navigation-factor data fusion method | |
| CN112364740B (en) | Unmanned aerial vehicle room monitoring method and system based on computer vision | |
| CN111428647A (en) | Traffic signal lamp fault detection method | |
| CN115082834B (en) | Engineering vehicle black smoke emission monitoring method and system based on deep learning | |
| CN111220619B (en) | Insulator self-explosion detection method | |
| CN112037251A (en) | Method for monitoring marine vessel exhaust emission by using smart phone | |
| CN115065798A (en) | Big data-based video analysis monitoring system | |
| CN114092903A (en) | Lane line marking method, lane line detection model determining method, lane line detection method and related equipment | |
| CN116363573A (en) | Transformer substation equipment state anomaly identification method and system | |
| CN111444916A (en) | License plate positioning and identifying method and system under unconstrained condition | |
| CN114332781A (en) | Intelligent license plate recognition method and system based on deep learning | |
| CN118038165A (en) | Video-based automatic identification algorithm for black smoke event of ship | |
| CN113591643A (en) | Underground vehicle station entering and exiting detection system and method based on computer vision | |
| CN115588207A (en) | Monitoring video date recognition method based on OCR | |
| CN115082832A (en) | Information identification method, device and storage medium | |
| CN113870620A (en) | Ship identification method for simultaneously starting multiple AIS (automatic identification system) devices |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |