CN114937218A - Electric power facility inspection system and method based on image recognition - Google Patents
Electric power facility inspection system and method based on image recognition Download PDFInfo
- Publication number
- CN114937218A CN114937218A CN202210859885.5A CN202210859885A CN114937218A CN 114937218 A CN114937218 A CN 114937218A CN 202210859885 A CN202210859885 A CN 202210859885A CN 114937218 A CN114937218 A CN 114937218A
- Authority
- CN
- China
- Prior art keywords
- unmanned aerial
- aerial vehicle
- night vision
- target object
- brightness change
- 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
- 238000000034 method Methods 0.000 title claims abstract description 14
- 238000007689 inspection Methods 0.000 title claims description 25
- 230000008859 change Effects 0.000 claims abstract description 55
- 230000004297 night vision Effects 0.000 claims abstract description 55
- 238000004891 communication Methods 0.000 claims description 10
- 238000000605 extraction Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 2
- 230000004397 blinking Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000006855 networking Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 230000004438 eyesight Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000001931 thermography Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/17—Terrestrial scenes taken from planes or by drones
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- Evolutionary Computation (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Remote Sensing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Water Supply & Treatment (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Image Analysis (AREA)
Abstract
The application discloses electric power facility system and method of patrolling and examining based on image recognition relates to image recognition and unmanned aerial vehicle technique, includes: an unmanned aerial vehicle carrying a night vision device and a camera; the special equipment is provided with a flash lamp; the unmanned aerial vehicle is used for shooting a night vision image through the night vision device, identifying a target object from the night vision image and broadcasting identification characteristic claim information; the special equipment receives the broadcast, then claims the identification characteristic and reports the identity information, and generates a control signal of the flash lamp according to the identification characteristic so as to control the flash frequency of the flash lamp; the unmanned aerial vehicle tracks and shoots the target object through the camera or the night vision device, and determines the identification characteristics according to the brightness change of the area where the target object is located so as to determine the identity information of the target object. The scheme can realize target distinguishing at night.
Description
Technical Field
The application relates to an image recognition technology and an unmanned aerial vehicle technology, in particular to an electric power facility inspection system and method based on image recognition.
Background
With the gradual maturity of unmanned aerial vehicle technology and image processing technology, some patrol and rescue tasks can be executed by carrying a camera by an unmanned aerial vehicle in many occasions. In some important facilities (such as power facilities) safeguard measures, unmanned aerial vehicles, robots and other inspection modes are not lacked.
When patrolling and examining in the daytime, because the environment has better light for unmanned aerial vehicle can carry out operations such as target identification, identification based on the image of shooing. However, at night, the camera cannot obtain good light at night, and therefore the definition of the shot image cannot meet the requirement of partial recognition. And carry on high-power light or high accuracy camera and can make unmanned aerial vehicle cost increase, have equipment to be difficult to load.
Although some schemes can be equipped with a night vision camera, the night vision camera can only be used for finding a target, and the identity of the target cannot be identified. When a plurality of targets are found in a patrol area, the identities of the targets cannot be distinguished, so that the tracked objects cannot be determined when the unmanned aerial vehicle needs to perform target tracking.
Disclosure of Invention
The present invention is directed to solving at least one of the problems of the prior art. Therefore, the invention provides an electric power facility inspection system and method based on image recognition, which can recognize the identity of a target at night based on an image recognition technology.
In one aspect, an embodiment of the present application provides an electric power facility inspection system based on image recognition, including:
an unmanned aerial vehicle carrying a night vision device and a camera;
the special equipment is provided with a flash lamp;
the unmanned aerial vehicle is used for shooting a night vision image through the night vision device, identifying a target object from the night vision image and broadcasting identification characteristic claim information;
the special equipment receives the broadcast, then claims the identification characteristics and reports the identity information, and generates a control signal of the flash lamp according to the identification characteristics so as to control the flash frequency of the flash lamp;
the unmanned aerial vehicle tracks and shoots the target object through the camera or the night vision device, and determines the identification characteristics according to the brightness change of the area where the target object is located so as to determine the identity information of the target object.
In some embodiments, the determining the identification feature according to the brightness change of the region where the target object is located specifically includes:
extracting picture frames in the shot video at intervals;
cutting out an area where a target is located from the picture frame to obtain a plurality of sub-pictures;
performing brightness statistics on the subgraph to determine a brightness change curve;
an identification feature is determined from the frequency of the waveform of the luminance profile.
In some embodiments, the drone determines the decimation interval of a picture frame according to the highest value of the flicker frequency among all the identification features claimed.
In some embodiments, the special device claims the identification feature and reports the identity information after receiving the broadcast, specifically:
after receiving the broadcast, the special equipment establishes communication with the unmanned aerial vehicle, reports the identity to the unmanned aerial vehicle, receives the identification feature number distributed by the unmanned aerial vehicle, and determines the identification feature according to the identification feature number.
In some embodiments, the drone is further configured to, prior to broadcasting the identifying feature claim information, further include the steps of:
acquiring the number of the special equipment in the current area through the Internet;
searching for a target object in a current area through a night vision device, and broadcasting identification feature claim information when the number of the target objects is greater than the number of the dedicated devices;
or,
searching target objects in the current area through a camera, and acquiring identity information corresponding to the special equipment in the current area through the Internet when the number of the target objects is 1.
In some embodiments, when the unmanned aerial vehicle determines that the distance between the two target objects is smaller than the threshold, the unmanned aerial vehicle filters the brightness change curves corresponding to the two target objects, so as to determine the brightness change curve of the target, and determines the identification feature based on the brightness change curve of the target.
In some embodiments, the luminance variation curve is band-pass filtered according to the frequency corresponding to the currently claimed identification feature, a candidate curve corresponding to each identification feature is separated, and the candidate curve with the maximum intensity is used as the target luminance variation curve.
In some embodiments, when the unmanned aerial vehicle determines the identification feature according to the brightness change of the area where the target object is located, a first brightness change curve corresponding to the night vision device and a second brightness change curve corresponding to the camera are respectively calculated;
and verifying the first brightness change curve by using the second brightness change curve, and determining the identification characteristics based on the first brightness change curve when the flicker frequencies corresponding to the first brightness change curve and the second change curve meet the condition.
In some embodiments, the dedicated device comprises a handset and a helmet, and the flashlight is disposed on the handset or the helmet.
An electric power facility inspection method based on image recognition is applied to an electric power facility inspection system based on image recognition, and comprises the following steps:
the unmanned aerial vehicle shoots a night vision image through the night vision device, identifies a target object from the night vision image, and broadcasts identification characteristic claim information;
the special equipment receives the broadcast, then claims the identification characteristic and reports the identity information, and generates a control signal of the flash lamp according to the identification characteristic so as to control the flash frequency of the flash lamp;
the unmanned aerial vehicle tracks and shoots the target object through the camera or the night vision device, and determines the identification characteristics according to the brightness change of the area where the target object is located so as to determine the identity information of the target object.
According to the embodiment of the application, the night vision device and the camera are mounted on the unmanned aerial vehicle, so that a target can be accurately identified at night, and operations such as tracking can be performed in places with sufficient light; the system configures special equipment for the inspection personnel, and the special equipment is provided with a flash lamp; the unmanned aerial vehicle can shoot a night vision image through the night vision device, identify a target object from the night vision image and then broadcast identification feature claim information; the special equipment receives the broadcast, then claims the identification characteristic and reports the identity information, and generates a control signal of the flash lamp according to the identification characteristic so as to control the flash frequency of the flash lamp; the unmanned aerial vehicle can track and shoot the target object through the camera or the night vision device, and determines the identification characteristics according to the brightness change of the area where the target object is located so as to determine the identity information of the target object; through this mode in remote area, the relatively poor region of ground locating signal cover can realize the identification at night, can determine field personnel's identity like this, distinguishes fast and patrols and examines personnel and invader etc. and flash of light signal is difficult to be forged simultaneously, under the condition with unmanned aerial vehicle communication negotiation discernment characteristic, can't hide through the mode of imitative flash light scintillation.
Drawings
In order to more clearly illustrate the technical solution in the embodiments of the present invention, a brief description will be given below of the drawings used in the description of the embodiments.
Fig. 1 is a block diagram of an electric power facility inspection system based on image recognition according to an embodiment of the present application;
FIG. 2 is a waveform diagram of a luminance variation curve provided by an embodiment of the present application;
fig. 3 is a flowchart of an electric power facility inspection method based on image recognition according to an embodiment of the present application.
Detailed Description
In order to make the purpose, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be clearly and completely described below through embodiments with reference to the accompanying drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It can be understood that unmanned aerial vehicle system of patrolling and examining can patrol and examine the important facility along the line in the condition such as night, monitors the personnel's condition in the important place, distinguishes whether the condition such as illegal break-in takes place. Meanwhile, the tasks such as night searching, target tracking and the like can be considered. The unmanned aerial vehicle inspection system is one of important means for inspecting huge electric power facilities. The image recognition technology is one of the important branches of the artificial intelligence technology, and analyzes a shot image by using technical means such as a neural network to recognize the position, type and the like of a target in the image.
Referring to fig. 1, the embodiment of the application discloses an electric power facility inspection system based on image recognition, including:
unmanned aerial vehicle carries on night-time vision device and camera. It is of course understood that the drone is used for processing means and communication means for data processing. Night vision devices may employ devices such as thermal imaging or infrared night vision devices, which are used primarily for the purpose of finding and tracking the location of objects during the night.
The special equipment is provided with a flash lamp. The dedicated device may be a handset, a dedicated helmet, or a combination of both. The flash lamp is arranged to send out an indication signal in the dark environment at night, so that the unmanned aerial vehicle in the air can identify the identity of the target in the picture based on the signal of the flash lamp. Taking the next step of the task. A dedicated device is generally equipped with a plurality of communication modules to cope with an outdoor complex communication environment.
The unmanned aerial vehicle is used for shooting a night vision image through the night vision device, identifying a target object from the night vision image and broadcasting identification characteristic claim information; it can be understood that the onboard night vision device can find a target at night, while the camera can perform a shooting task in a place with sufficient light or in the daytime, and meanwhile, the camera takes on the task of light identification in the embodiment. Of course, at night, the camera may also start recording. After the drone finds the target, the drone can broadcast the medium-short range in a frequency band receivable by the dedicated device, and the drone can broadcast in a frequency band such as 433MHz, 2.4GHz, and the range of the broadcast is generally controlled by the transmission power, and is generally controlled in a range from tens of meters to hundreds of meters. The method requires a special device to be capable of connecting with the server on site in a field broadcasting mode, and can adapt to a complex communication environment.
The special equipment receives the broadcast, then claims the identification characteristics and reports the identity information, and generates a control signal of the flash lamp according to the identification characteristics so as to control the flash frequency of the flash lamp.
Specifically, the dedicated device establishes communication with the unmanned aerial vehicle after receiving the broadcast, reports the identity to the unmanned aerial vehicle, receives the identification feature number assigned by the unmanned aerial vehicle, and determines the identification feature according to the identification feature number. It can be understood that, in some embodiments, the information about the relevant identification features, such as the flashing frequency and the flashing duty cycle, has been stored in the dedicated device in advance, and the drone may determine the identity of the target object based on the identity information reported by the dedicated device for the identification features only by negotiating which identification feature corresponds to which number is used between the dedicated device and the drone. As for the identity information in the dedicated device, it may be pre-bound, for example, a user of the handset logs in with the handset before the job, and at this time, the body code of the handset is bound with the identity information and uploaded to the server. At this point, the drone may access this information through the server. Of course, the task may be performed by downloading the data in advance in the case of off-line on site. Under the extreme condition communication circumstances, unmanned aerial vehicle and handheld machine are the line state, and unmanned aerial vehicle can record handheld machine's information, accomplishes the affirmation of identity after networking afterwards.
The unmanned aerial vehicle tracks and shoots the target object through the camera or the night vision device, and determines the identification characteristics according to the brightness change of the area where the target object is located so as to determine the identity information of the target object. It will be appreciated that when the flash of the dedicated device flashes, the emitted light will change the local brightness. This change is particularly apparent in the area of the target. When the flashlight is arranged on the safety helmet, the change can be relatively clear under the aerial photographing condition. Unmanned aerial vehicle passes through the continuous shooting of camera, then detects luminance change in the picture frame of video, just can determine the change law of flash light. At this time, the frequency of flash blinking and the duty ratio of blinking may be the identification characteristics. In the present embodiment, the intensity is not adopted as the identification feature because the brightness at night is uncertain and the intensity value cannot be detected without a reference value. Of course, in some embodiments, the flash may emit light of a specific color, and the light of the specific color may be filtered from the image frame for analysis, i.e., to analyze the brightness change of the light source of the specific color, before processing the brightness change curve. It can be understood that by adopting the above means, the accuracy is better at night compared with the mode of distinguishing identities based on different colors, and the device is not required to be provided with multicolor light; compared with a mode of controlling different devices to stagger and flicker, the method can identify the identities of a plurality of targets at the same time, and is higher in efficiency.
Specifically, the determining the identification feature according to the brightness change of the region where the target object is located specifically includes:
and extracting picture frames in the shot video at intervals. It is understood that the video may be a video captured by a night vision device or a video captured by a camera. In this embodiment, the target can be accurately identified and locked by the night vision device. Changes to the light can be captured in both the night vision device and the camera at night. Therefore, the frames extracted from the picture can be analyzed for the relevant flicker characteristics.
Cutting out the area where the target is located from the picture frame to obtain a plurality of sub-pictures; the region where the target needs to be cut out is characterized in that the flash lamp can only affect the brightness nearby the target, and by the mode, subgraphs corresponding to a plurality of target objects can be cut out in one picture, and the targets are identified simultaneously. For one target, corresponding subgraphs are intercepted from each extracted picture frame. Then the subgraphs form a subgraph sequence according to the time sequence. Fig. 2 shows a process of changing the brightness of the lamp, and it can be seen from the figure that, when the lamp flickers, the brightness of the frame of the picture changes according to the characteristics, and after a few cycles of acquisition, the waveform of the flash lamp can be analyzed through the corresponding time point of the frame of the picture, so as to distinguish the identity of the target object from the frequency characteristics of the waveform.
And carrying out brightness statistics on the subgraph to determine a brightness change curve. And counting the average brightness of each subgraph, and then drawing a brightness change curve according to the sequence of the subgraphs. Interpolation between data points can be performed to obtain a smooth curve. It can be understood that the recognition can be completed within 2-3 seconds, and in this time, the moving range of the general person is not large, and the brightness of the periphery is not obviously changed. Even if the user moves to cause the brightness of the environment to change obviously, the identity of the user can be determined in a continuous recognition mode. In the graph, the graph may be plotted with the average luminance of the first frame picture as a reference zero point.
An identification feature is determined from the frequency of the waveform of the luminance profile. It is understood that after obtaining the brightness variation curve, the flicker frequency can be obtained by measuring the distance between the peaks, and the flicker duty ratio can also be obtained by the related waveform processing algorithm. Then, the unmanned aerial vehicle can determine the identity of the identification object through the identification feature.
In some embodiments, in order to reduce the amount of image processing by the drone, the drone determines the extraction interval of the picture frames according to the highest value of the flicker frequency among all the identification features claimed. It will be appreciated that in order to maintain the accuracy of the measurement, it is necessary to ensure that at least 10 frames of the picture are extracted for processing each flash cycle, and then for 60Hz recording, it is necessary to extract one frame every 6 frames, i.e. every 5 frames. Then, assuming that three objects are currently identified, which are respectively flashed 1 time per second, 2 times per second and 1 time per 2 seconds, at this time, the highest value of the flashing frequency is 2 times per second, which means that 20 frames of pictures need to be extracted for processing within 1 second in order to acquire at least 10 frames of pictures per period, and for a video recording of 60Hz, one frame is extracted every three frames. Through this mode, for setting for fixed extraction interval, can ensure to reduce the handling capacity under the condition of discernment quality at will, save unmanned aerial vehicle consumption. It should be understood that, since the drone does not know which flicker frequency the object currently being analyzed corresponds to specifically before performing the analysis, setting the extraction interval based on the highest frequency can ensure that the number of frames extracted per flicker period is not less than the set minimum value.
In some embodiments, the drone is further configured to, prior to broadcasting the identifying feature claim information, further include the steps of:
the number of dedicated devices in the current area is obtained via the internet. It will be appreciated that where networking is possible, the drone may communicate with the server over the internet to determine the number of people (number of devices) that are performing the task in the current area.
Searching for a target object in a current area through a night vision device, and broadcasting identification feature claim information when the number of target objects is greater than the number of dedicated devices. In order to reduce the identification pressure, the drone may broadcast the identification feature claim only when searching for a target object in the current area that is larger than the number of dedicated devices. Therefore, the number of identification times can be reduced, and the links can be saved on the occasions partially without distinguishing the identities of the on-site personnel.
Or,
searching target objects in the current area through a camera, and acquiring identity information corresponding to the special equipment in the current area through the Internet when the number of the target objects is 1. In some embodiments, if the current region has only one target object and the identity of the current worker can be determined based on the internet, no distinction between workers may be required, which may reduce the power consumption of the device resulting from recognition.
In some embodiments, when two targets are closer, the problem of mutual influence of the flashes of the dedicated devices of the two targets needs to be considered, but because the two targets are still located at different distances from the flashes, the brightness influence intensities of the two flashes on the two targets are different. In this embodiment, because the unmanned aerial vehicle knows each flicker frequency, the flicker frequency difference is usually large, so two waves with different frequencies can be separated from the brightness change curve in a filtering manner, and then the intensity influence caused by the distance is different, and the frequency of the light source close to the target object can be determined, so that the mutual influence when two flash lamps flicker simultaneously is avoided. It will be appreciated that the distance between two targets is estimated based on the distance of the drone from the target. The size of the target shot by the unmanned aerial vehicle in the picture is related to the distance between the unmanned aerial vehicle and the target. Therefore, a simple determination method can be adopted, namely whether the distance between the two target frames of the target object in the picture is smaller than the threshold value or not is determined according to the ratio of the distance between the two target frames and the size of the target frames. Generally, the height of the person is not very different, and therefore, the sizes of two similar object boxes are close. The target frame is large, which shows that the target is relatively close to the unmanned aerial vehicle; the target frames are smaller, which means that the targets are relatively far from the drone, and when the distance between two sets of target frames is the same, it is obvious that the distance between one set of target frames is larger. Thus, the distance between two target frames may be divided by the average size of the target frames to obtain a ratio value, and whether the distance is small may be determined according to whether the ratio value is smaller than a threshold value. In this way, it is possible to directly estimate whether the distance between two target objects is too small in the screen based on the empirical value without performing accurate measurement. Therefore, when the drone determines that the distance between the two target objects is smaller than the threshold, specifically, when the drone determines that the distance between the target frames of the two target objects in the screen divided by the area of one of the target frames (which may be the one with the larger area) is smaller than the threshold, the drone determines that the distance between the two target objects is smaller than the threshold. It is to be understood that the target frame is a target area that is framed when the image recognition model recognizes that the target is selected.
In some embodiments, the luminance variation curve is band-pass filtered according to the frequency corresponding to the currently claimed identification feature, a candidate curve corresponding to each identification feature is separated, and the candidate curve with the maximum intensity (which may be measured by an average amplitude) is used as the luminance variation curve of the target. It is understood that the influence of the flash near the target object on the brightness of the region where the target object is located is greater. Therefore, the brightness variation curve separated from the target object is stronger than the light emitted by the special equipment held by the target object. Based on the embodiment, the identities of the two targets can be distinguished under the condition that the two target objects are close and flicker simultaneously. In this way, there is no need to further identify the flash location to improve accuracy, so the algorithm is simple and reliable.
In some embodiments, when the unmanned aerial vehicle determines the identification feature according to the brightness change of the area where the target object is located, a first brightness change curve corresponding to the night vision device and a second brightness change curve corresponding to the camera are respectively calculated;
and verifying the first brightness change curve by using the second brightness change curve, and determining the identification characteristic based on the first brightness change curve when the flicker frequency corresponding to the first brightness change curve and the second brightness change curve meets the condition. In some embodiments, the calculated recognition features may not be accurate due to frame loss, and in this case, the pictures taken by the two cameras may be used for mutual calibration in the case of sufficient calculation. The frequencies of the luminance variation curves corresponding to the two are mainly calibrated.
In some embodiments, the dedicated device comprises a handset and a helmet, and the flashlight is disposed on the handset or the helmet. It will be appreciated that the flash may be provided on a helmet, the flash on the helmet communicating with the handset, under the control of the handset. It can be understood that general operation personnel need wear the helmet to set up the flash light on the helmet, unmanned aerial vehicle can clearly discern light signal night.
Referring to fig. 3, the embodiment discloses an electric power facility inspection method based on image recognition, which is applied to an electric power facility inspection system based on image recognition, and includes:
s1, shooting a night vision image by the unmanned aerial vehicle through the night vision device, identifying a target object from the night vision image, and broadcasting identification characteristic claim information;
s2, the special device claims the identification characteristics and reports the identity information after receiving the broadcast, and generates the control signal of the flash lamp according to the identification characteristics to control the flash frequency of the flash lamp;
and S3, the unmanned aerial vehicle tracks and shoots the target object through the camera or the night vision device, and determines the identification characteristics according to the brightness change of the area where the target object is located so as to determine the identity information of the target object.
In the embodiment, the night vision device and the camera are arranged on the unmanned aerial vehicle, so that the target can be accurately identified at night, and operations such as tracking can be performed in places with sufficient light; the system configures special equipment for inspection personnel, and the special equipment is provided with a flash lamp; the unmanned aerial vehicle can shoot a night vision image through the night vision device, identify a target object from the night vision image, and then broadcast identification feature claim information; the special equipment receives the broadcast, then claims the identification characteristic and reports the identity information, and generates a control signal of the flash lamp according to the identification characteristic so as to control the flash frequency of the flash lamp; the unmanned aerial vehicle can track and shoot the target object through the camera or the night vision device, and determines the identification characteristics according to the brightness change of the area where the target object is located so as to determine the identity information of the target object; through this mode in remote area, the relatively poor region of ground locating signal cover can realize the identification at night, can determine field personnel's identity like this, distinguishes fast and patrols and examines personnel and invader etc. and flash of light signal is difficult to be forged simultaneously, under the condition with unmanned aerial vehicle communication negotiation discernment characteristic, can't hide through the mode of imitative flash light scintillation.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present application and the technical principles employed. It will be understood by those skilled in the art that the present application is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the application. Therefore, although the present application has been described in more detail with reference to the above embodiments, the present application is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present application, and the scope of the present application is determined by the scope of the appended claims.
Claims (10)
1. The utility model provides an electric power facility system of patrolling and examining based on image recognition which characterized in that includes:
an unmanned aerial vehicle carrying a night vision device and a camera;
the special equipment is provided with a flash lamp;
the unmanned aerial vehicle is used for shooting a night vision image through the night vision device, identifying a target object from the night vision image and broadcasting identification characteristic claim information;
the special equipment receives the broadcast, then claims the identification characteristics and reports the identity information, and generates a control signal of the flash lamp according to the identification characteristics so as to control the flash frequency of the flash lamp;
the unmanned aerial vehicle tracks and shoots the target object through the camera or the night vision device, and determines the identification characteristics according to the brightness change of the area where the target object is located so as to determine the identity information of the target object.
2. The image recognition-based power facility inspection system according to claim 1, wherein the determining of the recognition characteristics according to the brightness change of the area where the target object is located specifically comprises:
extracting picture frames in the shot video at intervals;
cutting out the area where the target is located from the picture frame to obtain a plurality of sub-pictures;
performing brightness statistics on the subgraph to determine a brightness change curve;
an identification feature is determined from the frequency of the waveform of the brightness variation curve.
3. The image recognition-based electric power facility inspection system according to claim 2, wherein the unmanned aerial vehicle determines the extraction interval of the picture frames according to the highest value of the flicker frequency among all the recognition features claimed, wherein the number of frames extracted per flicker period is set to the minimum value.
4. The image recognition-based power facility inspection system according to claim 1, wherein the special equipment claims recognition features and reports identity information after receiving the broadcast, and specifically comprises:
after receiving the broadcast, the special equipment establishes communication with the unmanned aerial vehicle, reports the identity to the unmanned aerial vehicle, receives the identification feature number distributed by the unmanned aerial vehicle, and determines the identification feature according to the identification feature number.
5. The image recognition-based power facility inspection system according to claim 1, wherein the drone is further configured to, prior to broadcasting the identification feature claim information, further include:
acquiring the number of the special equipment in the current area through the Internet;
searching for a target object in a current area through a night vision device, and broadcasting identification feature claim information when the number of the target objects is greater than the number of the dedicated devices;
or,
searching target objects in the current area through a camera, and acquiring identity information corresponding to the special equipment in the current area through the Internet when the number of the target objects is 1.
6. The image recognition-based power facility inspection system according to claim 1, wherein the unmanned aerial vehicle filters brightness change curves corresponding to two target objects when determining that the distance between the two target objects is smaller than a threshold value, so as to determine the brightness change curves of the targets, and determines the recognition features based on the brightness change curves of the targets.
7. The image recognition-based power facility inspection system according to claim 6, wherein the candidate curves corresponding to the recognition features are separated by band-pass filtering the brightness variation curve according to the frequency corresponding to the currently claimed recognition feature, and the candidate curve with the highest intensity is used as the target brightness variation curve.
8. The image recognition-based power facility inspection system according to claim 1, wherein the unmanned aerial vehicle respectively calculates a first brightness change curve corresponding to the night vision device and a second brightness change curve corresponding to the camera when determining the recognition characteristics according to the brightness change of the area where the target object is located;
and verifying the first brightness change curve by using the second brightness change curve, and determining the identification characteristics based on the first brightness change curve when the flicker frequencies corresponding to the first brightness change curve and the second change curve meet the condition.
9. The image recognition-based electric power facility inspection system according to claim 1, wherein the dedicated device includes a handset and a helmet, and the flash is provided on the handset or the helmet.
10. An electric power facility inspection method based on image recognition, which is applied to the system according to claim 1, and comprises the following steps:
the unmanned aerial vehicle shoots a night vision image through the night vision device, identifies a target object from the night vision image, and broadcasts identification characteristic claim information;
the special equipment claims the identification characteristics and reports the identity information after receiving the broadcast, and generates a control signal of the flash lamp according to the identification characteristics so as to control the flash frequency of the flash lamp;
the unmanned aerial vehicle tracks and shoots the target object through the camera or the night vision device, and determines the identification characteristics according to the brightness change of the area where the target object is located so as to determine the identity information of the target object.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210859885.5A CN114937218B (en) | 2022-07-22 | 2022-07-22 | Electric power facility inspection system and method based on image recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210859885.5A CN114937218B (en) | 2022-07-22 | 2022-07-22 | Electric power facility inspection system and method based on image recognition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114937218A true CN114937218A (en) | 2022-08-23 |
| CN114937218B CN114937218B (en) | 2022-10-21 |
Family
ID=82868444
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210859885.5A Active CN114937218B (en) | 2022-07-22 | 2022-07-22 | Electric power facility inspection system and method based on image recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114937218B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116027798A (en) * | 2022-09-30 | 2023-04-28 | 三峡大学 | Unmanned aerial vehicle power inspection system and method based on image correction |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020063702A1 (en) * | 2000-11-30 | 2002-05-30 | Nec Corporation | Dynamic image processing method, dynamic image processor and dynamic image display device |
| JP2006253008A (en) * | 2005-03-11 | 2006-09-21 | Matsushita Electric Ind Co Ltd | Flicker suppressing device of discharge lamp |
| US20110305381A1 (en) * | 2010-06-10 | 2011-12-15 | Yoshinori Ohno | Radiological image area extracting apparatus, radiological image area extraction program, and radiographic apparatus |
| CN103327179A (en) * | 2013-06-03 | 2013-09-25 | 深圳市宝尔爱迪科技有限公司 | System and implementation method for mobile phone lamp signal transmission and recognition |
| US20140300475A1 (en) * | 2013-03-15 | 2014-10-09 | Juan Enrique Cienfuegos | Method of utilizing a disaster ID system utilizing illuminated displays to communicate with disaster relief personnel |
| WO2015098108A1 (en) * | 2013-12-27 | 2015-07-02 | パナソニック インテレクチュアル プロパティ コーポレーション オブ アメリカ | Information processing program, receiving program and information processing device |
| CN206058207U (en) * | 2016-07-22 | 2017-03-29 | 浙江钧普科技股份有限公司 | Wireless police personnel identity recognizes instruction device |
| CN107040888A (en) * | 2017-02-27 | 2017-08-11 | 集美大学 | Watercraft identification automatic identification and law enforcement cruising inspection system and method based on radio frequency |
| CN109274984A (en) * | 2018-10-16 | 2019-01-25 | 深圳开阳电子股份有限公司 | Detect the method, apparatus and image processing equipment of light source scintillation in video sequence |
| CN113810632A (en) * | 2021-10-19 | 2021-12-17 | 深圳市迈测科技股份有限公司 | Infrared imaging device and night vision device |
| CN114424521A (en) * | 2019-10-11 | 2022-04-29 | 谷歌有限责任公司 | Reducing flicker effects of multiple light sources in an image |
| CN114580823A (en) * | 2021-11-23 | 2022-06-03 | 国网甘肃省电力公司 | A Human-Machine Natural Interaction Method Based on Multi-mode Integration of Substations |
-
2022
- 2022-07-22 CN CN202210859885.5A patent/CN114937218B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020063702A1 (en) * | 2000-11-30 | 2002-05-30 | Nec Corporation | Dynamic image processing method, dynamic image processor and dynamic image display device |
| JP2006253008A (en) * | 2005-03-11 | 2006-09-21 | Matsushita Electric Ind Co Ltd | Flicker suppressing device of discharge lamp |
| US20110305381A1 (en) * | 2010-06-10 | 2011-12-15 | Yoshinori Ohno | Radiological image area extracting apparatus, radiological image area extraction program, and radiographic apparatus |
| US20140300475A1 (en) * | 2013-03-15 | 2014-10-09 | Juan Enrique Cienfuegos | Method of utilizing a disaster ID system utilizing illuminated displays to communicate with disaster relief personnel |
| CN103327179A (en) * | 2013-06-03 | 2013-09-25 | 深圳市宝尔爱迪科技有限公司 | System and implementation method for mobile phone lamp signal transmission and recognition |
| WO2015098108A1 (en) * | 2013-12-27 | 2015-07-02 | パナソニック インテレクチュアル プロパティ コーポレーション オブ アメリカ | Information processing program, receiving program and information processing device |
| CN206058207U (en) * | 2016-07-22 | 2017-03-29 | 浙江钧普科技股份有限公司 | Wireless police personnel identity recognizes instruction device |
| CN107040888A (en) * | 2017-02-27 | 2017-08-11 | 集美大学 | Watercraft identification automatic identification and law enforcement cruising inspection system and method based on radio frequency |
| CN109274984A (en) * | 2018-10-16 | 2019-01-25 | 深圳开阳电子股份有限公司 | Detect the method, apparatus and image processing equipment of light source scintillation in video sequence |
| CN114424521A (en) * | 2019-10-11 | 2022-04-29 | 谷歌有限责任公司 | Reducing flicker effects of multiple light sources in an image |
| CN113810632A (en) * | 2021-10-19 | 2021-12-17 | 深圳市迈测科技股份有限公司 | Infrared imaging device and night vision device |
| CN114580823A (en) * | 2021-11-23 | 2022-06-03 | 国网甘肃省电力公司 | A Human-Machine Natural Interaction Method Based on Multi-mode Integration of Substations |
Non-Patent Citations (3)
| Title |
|---|
| HUNG M.LA ET AL.: "development of an autonomous bridge deck inspection robotic system", 《JOURNAL OF FIELD ROBOTICS》 * |
| 张继荣: "基于无人机热红外遥感的地面人员检测模型研究", 《中国优秀硕士学位论文全文数据库 基础科学辑》 * |
| 石玗等: "基于激光视觉的脉冲GTAW熔池振荡检测与分析", 《机械工程学报》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116027798A (en) * | 2022-09-30 | 2023-04-28 | 三峡大学 | Unmanned aerial vehicle power inspection system and method based on image correction |
| CN116027798B (en) * | 2022-09-30 | 2023-11-17 | 三峡大学 | UAV power inspection system and method based on image correction |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114937218B (en) | 2022-10-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020199538A1 (en) | Bridge key component disease early-warning system and method based on image monitoring data | |
| CN100449576C (en) | Real-time image monitoring and genuine-fake identification system for vehicle with automatic color regulating and light filling license plate | |
| US20160260306A1 (en) | Method and device for automated early detection of forest fires by means of optical detection of smoke clouds | |
| CN111710055A (en) | Portable power inspection equipment, power inspection method and power inspection system | |
| CN104698427B (en) | Automatic checking method for faults of outdoor electric energy meter | |
| CN112396658A (en) | Indoor personnel positioning method and positioning system based on video | |
| CN112257554B (en) | Forest fire recognition method, system, program and storage medium based on multiple spectra | |
| CN102117484B (en) | Processing system, processing method and image classification method using image color information | |
| CN114937218B (en) | Electric power facility inspection system and method based on image recognition | |
| CN104316538A (en) | Flickering seam detection method and device for cable cladding process | |
| CN113095114A (en) | Method, device and equipment for detecting oil leakage of hydraulic system | |
| CN111695512A (en) | Unattended cultural relic monitoring method and device | |
| CN111862621B (en) | Intelligent snapshot system of multi-type adaptive black cigarette vehicle | |
| CN115953726B (en) | Machine vision container face damage detection method and system | |
| CN113778091A (en) | A kind of inspection method for wind farm booster station equipment | |
| CN109828891B (en) | Fault indicator lamp identification method | |
| KR101450119B1 (en) | Method for controlling light module of camera and apparatus thereof | |
| CN114878005A (en) | Temperature detection method, device, system, electronic equipment and storage medium | |
| CN111724605A (en) | Black smoke vehicle monitoring system and monitoring method thereof | |
| CN115346377B (en) | AVP vehicle detection system and method | |
| KR101829274B1 (en) | System and apparatus for image security using beacon | |
| CN215298410U (en) | Invasion identification system | |
| KR102624333B1 (en) | Deep Learning-based Fire Monitoring System | |
| CN115984898A (en) | Object detection method, object detection device, storage medium, and electronic device | |
| KR20190090100A (en) | Solar module monitoring system using flight path generation algorithm and monitoring method of solar module using it |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |