1. A safety precaution method, comprising:

pre-scanning the transformer substation with first preset precision to obtain a pre-scanning model;

performing correction scanning with second preset precision on the transformer substation to obtain a correction scanning model; obtaining a rotation value and a translation value according to the pre-scanning model and the correction scanning model; wherein the accuracy of the pre-scan is higher than the accuracy of the correction scan;

carrying out subsequent multiple recognition scanning with second preset precision on the transformer substation in real time, and correcting a recognition scanning model obtained by each recognition scanning according to the rotation value and the translation value to obtain a corrected recognition model;

carrying out differential operation on the correction recognition model and the pre-scanning model to obtain a target object; calculating the real-time distance from the target object to a cable in the pre-scanning model;

and judging whether to perform safety early warning on the target object according to the real-time distance.

2. The method of claim 1, wherein the determining whether to perform the safety precaution on the target object according to the real-time distance comprises:

identifying an object class of the target object;

acquiring a safe distance of the object category;

and comparing the real-time distance with the safe distance, and judging whether to perform safety early warning on the target object.

3. The method of claim 1, wherein obtaining rotation and translation values from the pre-scan model and the calibration scan model comprises:

selecting a plurality of corresponding pairs of key points from the pre-scanning model and the correction scanning model respectively;

and correcting the correction scanning model by matching each pair of the key points to obtain the rotation value and the translation value.

4. The method of claim 3, wherein after correcting the corrected scan model by matching each pair of the keypoints and obtaining the rotation value and the translation value, comprising:

extracting a region scanning model within a certain region correction scanning model range and a preset distance nearby the region scanning model range in the pre-scanning model; down-sampling the area scanning model and extracting the outer contour of the area scanning model after down-sampling;

and performing second preset precision quasi-correction scanning on the certain region again according to preset time according to the down-sampled outer contour of the region scanning model, correcting the outer contour of the extracted quasi-correction scanning model, and updating the rotation value and the translation value.

5. The method according to claim 4, wherein the performing, according to the downsampled outline of the area scan model, a second preset precision quasi-calibration scan on the certain area again according to a preset time, and the correcting the extracted outline of the quasi-calibration scan model and updating the rotation value and the translation value includes:

performing quasi-correction scanning with second preset precision on the certain area again according to preset time, and extracting to obtain the outer contour of the quasi-correction scanning model;

judging whether the difference value between the outer contour of the quasi-correction scanning model and the outer contour of the correction scanning model exceeds a preset threshold value or not; if not, automatically correcting the quasi-correction scanning model, performing down-sampling on the outer contour of the corrected quasi-correction scanning model, and automatically correcting the outer contour of the area scanning model after down-sampling; and if so, manually correcting the quasi-correction scanning model.

6. A safety precaution device, comprising:

the pre-scanning unit is used for pre-scanning the transformer substation with a first preset precision to obtain a pre-scanning model;

the correction scanning unit is used for performing correction scanning with second preset precision on the transformer substation to obtain a correction scanning model; obtaining a rotation value and a translation value according to the pre-scanning model and the correction scanning model; wherein the accuracy of the pre-scan is higher than the accuracy of the correction scan;

the correction identification unit is used for carrying out subsequent multiple recognition scanning with second preset precision on the transformer substation in real time, and correcting a recognition scanning model obtained by each recognition scanning according to the rotation value and the translation value to obtain a correction recognition model;

the calculation unit is used for carrying out differential operation on the correction recognition model and the pre-scanning model to obtain a target object; calculating the real-time distance from the target object to a cable in the pre-scanning model;

and the safety early warning unit is used for judging whether to carry out safety early warning on the target object according to the real-time distance.

7. The apparatus of claim 6, wherein the safety precaution unit comprises:

the identification subunit is used for identifying the object type of the target object;

the acquisition subunit is used for acquiring the safe distance of the object type;

and the safety early warning subunit is used for comparing the real-time distance with the safety distance and judging whether to carry out safety early warning on the target object.

8. The apparatus of claim 6, wherein the correction scanning unit comprises:

the selecting subunit is used for respectively selecting a plurality of corresponding pairs of key points from the pre-scanning model and the correction scanning model;

and the matching subunit is used for correcting the correction scanning model by matching each pair of the key points and obtaining the rotation value and the translation value.

9. The apparatus of claim 8, wherein the calibration scan unit further comprises:

the extraction subunit is used for extracting a region scanning model within a region correction scanning model range in the pre-scanning model and within a preset distance nearby the region correction scanning model range; down-sampling the area scanning model and extracting the outer contour of the area scanning model after down-sampling;

and the corrector subunit is used for performing quasi-correction scanning with second preset precision on the certain region again according to the down-sampled outer contour of the region scanning model and preset time, correcting the extracted outer contour of the quasi-correction scanning model, and updating the rotation value and the translation value.

10. A safety precaution system, comprising: the safety precaution device of any one of claims 6 to 9.

11. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the safety precaution method of any one of claims 1 to 5.

Background

For monitoring the safety distance of the cable in the transformer substation, the existing technology is mainly applied to the obstacle avoidance of the cable when the unmanned aerial vehicle is controlled to detect. Under this sight, only record the height of each position point cable in the meticulous model, through the height of contrast unmanned aerial vehicle current position to obtain the height of the relative cable of unmanned aerial vehicle.

But people, vehicles and animals entering the field can not carry the height sensor with the unmanned aerial vehicle. The position information of people, vehicles and animals in the space needs to be detected in real time through external equipment. This results in the need for accurate and fast detection equipment when determining the location of a foreign object in the substation.

The accuracy is that the cable is usually thin and requires high accuracy for position calculation. The safety distance between the foreign object and the cable is often between several meters, and the potential safety hazard can be generated by only a few centimeters. The method is fast in speed, people, vehicles and animals move fast and movement is uncontrollable, the unmanned aerial vehicle can not obtain instructions to act again, and if real-time judgment cannot be achieved, early warning cannot be timely carried out before safety problems occur to avoid the safety problems.

On the other hand, the accuracy and the rapidness are opposite, the laser precision of the rapid scanning is not high, and the laser speed with enough precision is not high.

The method aims at solving the problems that in the prior art, the position of a foreign object in a transformer substation is judged so as to ensure that the foreign object can be avoided by early warning in time before a safety problem occurs, and the accuracy and real-time cannot be ensured at the same time, and an effective solution is not provided at present.

Disclosure of Invention

The embodiment of the invention provides a safety early warning method, a safety early warning device, a safety early warning system and a computer storage medium, and aims to solve the problems that in the prior art, the position of a foreign object in a transformer substation is judged to ensure that the foreign object can be early warned in time before a safety problem occurs, and the accuracy and the real-time cannot be ensured at the same time.

In order to achieve the above object, in one aspect, the present invention provides a safety precaution method, including: pre-scanning the transformer substation with first preset precision to obtain a pre-scanning model; performing correction scanning with second preset precision on the transformer substation to obtain a correction scanning model; obtaining a rotation value and a translation value according to the pre-scanning model and the correction scanning model; wherein the accuracy of the pre-scan is higher than the accuracy of the correction scan; carrying out subsequent multiple recognition scanning with second preset precision on the transformer substation in real time, and correcting a recognition scanning model obtained by each recognition scanning according to the rotation value and the translation value to obtain a corrected recognition model; carrying out differential operation on the correction recognition model and the pre-scanning model to obtain a target object; calculating the real-time distance from the target object to a cable in the pre-scanning model; and judging whether to perform safety early warning on the target object according to the real-time distance.

Optionally, the determining whether to perform safety precaution on the target object according to the real-time distance includes: identifying an object class of the target object; acquiring a safe distance of the object category; and comparing the real-time distance with the safe distance, and judging whether to perform safety early warning on the target object.

Optionally, the obtaining a rotation value and a translation value according to the pre-scanning model and the calibration scanning model includes: selecting a plurality of corresponding pairs of key points from the pre-scanning model and the correction scanning model respectively; and correcting the correction scanning model by matching each pair of the key points to obtain the rotation value and the translation value.

Optionally, after the correcting the scan model by matching each pair of the keypoints, and obtaining the rotation value and the translation value, the method includes: extracting a region scanning model within a certain region correction scanning model range and a preset distance nearby the region scanning model range in the pre-scanning model; down-sampling the area scanning model and extracting the outer contour of the area scanning model after down-sampling; and performing second preset precision quasi-correction scanning on the certain region again according to preset time according to the down-sampled outer contour of the region scanning model, correcting the outer contour of the extracted quasi-correction scanning model, and updating the rotation value and the translation value.

Optionally, the performing, according to the downsampled outline of the area scanning model, second preset-precision quasi-correction scanning on the certain area again according to preset time, and the extracting the outline of the quasi-correction scanning model to correct and update the rotation value and the translation value includes: performing quasi-correction scanning with second preset precision on the certain area again according to preset time, and extracting to obtain the outer contour of the quasi-correction scanning model; judging whether the difference value between the outer contour of the quasi-correction scanning model and the outer contour of the correction scanning model exceeds a preset threshold value or not; if not, automatically correcting the quasi-correction scanning model, performing down-sampling on the outer contour of the corrected quasi-correction scanning model, and automatically correcting the outer contour of the area scanning model after down-sampling; and if so, manually correcting the quasi-correction scanning model.

In another aspect, the present invention provides a safety precaution device, including: the pre-scanning unit is used for pre-scanning the transformer substation with a first preset precision to obtain a pre-scanning model; the correction scanning unit is used for performing correction scanning with second preset precision on the transformer substation to obtain a correction scanning model; obtaining a rotation value and a translation value according to the pre-scanning model and the correction scanning model; wherein the accuracy of the pre-scan is higher than the accuracy of the correction scan; the correction identification unit is used for carrying out subsequent multiple recognition scanning with second preset precision on the transformer substation in real time, and correcting a recognition scanning model obtained by each recognition scanning according to the rotation value and the translation value to obtain a correction recognition model; the calculation unit is used for carrying out differential operation on the correction recognition model and the pre-scanning model to obtain a target object; calculating the real-time distance from the target object to a cable in the pre-scanning model; and the safety early warning unit is used for judging whether to carry out safety early warning on the target object according to the real-time distance.

Optionally, the safety precaution unit includes: the identification subunit is used for identifying the object type of the target object; the acquisition subunit is used for acquiring the safe distance of the object type; and the safety early warning subunit is used for comparing the real-time distance with the safety distance and judging whether to carry out safety early warning on the target object.

Optionally, the calibration scanning unit includes: the selecting subunit is used for respectively selecting a plurality of corresponding pairs of key points from the pre-scanning model and the correction scanning model; and the matching subunit is used for correcting the correction scanning model by matching each pair of the key points and obtaining the rotation value and the translation value.

Optionally, the calibration scanning unit further includes: the extraction subunit is used for extracting a region scanning model within a region correction scanning model range in the pre-scanning model and within a preset distance nearby the region correction scanning model range; down-sampling the area scanning model and extracting the outer contour of the area scanning model after down-sampling; and the corrector subunit is used for performing quasi-correction scanning with second preset precision on the certain region again according to the down-sampled outer contour of the region scanning model and preset time, correcting the extracted outer contour of the quasi-correction scanning model, and updating the rotation value and the translation value.

On the other hand, the invention also provides a safety early warning system which comprises the safety early warning device.

In another aspect, the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program is executed by a processor to implement the above-mentioned safety precaution method.

The invention has the beneficial effects that:

the invention provides a safety early warning method, which comprises the steps of scanning a transformer substation with a first preset precision and a second preset precision to obtain a pre-scanning model, a correction scanning model and a correction identification model, and judging the positions of human, vehicle, animal and other foreign objects in the transformer substation to ensure that the foreign objects can be early warned in time before safety problems occur, so that the problems of accuracy and real time are simultaneously ensured; and obtaining a rotation value and a translation value according to the pre-scanning model and the correction scanning model, performing subsequent recognition scanning with second preset precision on the transformer substation for multiple times, and directly correcting the recognition scanning model obtained by each recognition scanning according to the rotation value and the translation value to obtain a correction recognition model.

Drawings

Fig. 1 is a flowchart of a safety precaution method according to an embodiment of the present invention;

fig. 2 is a flowchart illustrating a process of determining whether to perform a safety warning on a target object according to a real-time distance according to an embodiment of the present invention;

FIG. 3 is a flowchart for obtaining rotation values and translation values according to a pre-scan model and a calibration scan model according to an embodiment of the present invention;

fig. 4 is a schematic structural diagram of a safety precaution device provided in an embodiment of the present invention;

fig. 5 is a schematic structural diagram of a safety precaution unit provided in an embodiment of the present invention;

fig. 6 is a schematic structural diagram of a calibration scan unit according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention clearer, the present invention will be described in further detail with reference to the accompanying drawings, and it is apparent that the described embodiments are only a part of the embodiments of the present invention, not all of the 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 invention.

For monitoring the safety distance of the cable in the transformer substation, the existing technology is mainly applied to the obstacle avoidance of the cable when the unmanned aerial vehicle is controlled to detect. Under this sight, only record the height of each position point cable in the meticulous model, through the height of contrast unmanned aerial vehicle current position to obtain the height of the relative cable of unmanned aerial vehicle.

But people, vehicles and animals entering the field can not carry the height sensor with the unmanned aerial vehicle. The position information of people, vehicles and animals in the space needs to be detected in real time through external equipment. This leads to the problem that it is not possible to ensure both accuracy and real-time when determining the location of a foreign object in the substation.

Therefore, the present invention provides a safety precaution method, fig. 1 is a flowchart of a safety precaution method provided in an embodiment of the present invention, and as shown in fig. 1, the method includes:

s101, pre-scanning the transformer substation with first preset precision to obtain a pre-scanning model;

s102, performing correction scanning with second preset precision on the transformer substation to obtain a correction scanning model; obtaining a rotation value and an offset value according to the pre-scanning model and the correction scanning model; wherein the accuracy of the pre-scan is higher than the accuracy of the correction scan;

in an optional embodiment, the first preset precision is 1mm, the second preset precision is 3mm, the transformer substation scene is finely scanned by using equipment with the scanning precision of 1mm to obtain a pre-scanning model, and the transformer substation scene is roughly scanned once by using equipment with the scanning precision of 3mm to obtain a correction scanning model; the cables in the pre-scanning model are very clear, and the correction scanning model is influenced by the scanning precision, so that the obtained cables have breakpoints and covering conditions;

at this time, the corrected scanning model and the pre-scanning model are corrected to obtain a rotation value and a translation value at the same time, and the obtained rotation value and translation value are used for subsequent calculation.

S103, carrying out subsequent multiple recognition scanning with second preset precision on the transformer substation in real time, and correcting a recognition scanning model obtained by each recognition scanning according to the rotation value and the translation value to obtain a corrected recognition model;

in an optional embodiment, the substation is subjected to subsequent multiple recognition scanning with second preset precision by using equipment with the scanning precision of 3 mm; for example, the subsequent multiple recognition scans are … … of 2 nd, 3 th, and 4 … … 99 th, and the recognition scan model obtained from each recognition scan is directly corrected according to the above rotation value and translation value to obtain a corrected recognition model, specifically, the point P (X, Y, Z) on the rough scan model is transformed according to the following formula to obtain the fine scan model P (X, Y, Z):

wherein (X, Y, Z) is a point on the rough scan model, (X, Y, Z) is a point on the fine scan model, R is the rotation value, and T is the translation value.

Furthermore, the interval time of every two adjacent identification scans is the same, and two identification scans can be realized within one second; assuming that there are 99 of the corrected recognition models; in the above 99 identification scans, there may be foreign objects such as people, cars, animals, etc., for example, a person appears in the 2 nd identification scan, a person and a car appear in the 20 th identification scan, a person and two cars appear in the 50 th identification scan, and a plurality of persons, a plurality of cars and a plurality of animals appear in the 99 th identification scan. In the process, each subsequent identification scanning is directly corrected according to the rotation value R and the translation value T obtained by the first correction, the rotation value and the translation value do not need to be recalculated, the correction time of operators is saved, and the correction speed is accelerated.

S104, carrying out differential operation on the correction recognition model and the pre-scanning model to obtain a target object; calculating the real-time distance from the target object to a cable in the pre-scanning model;

in an optional embodiment, the calibration identification model and the pre-scan model are subjected to a differential operation to remove a background in the substation, so as to obtain the target object, for example: humans, cars, animals, etc.; calculating the real-time distance from the target object to a cable in the pre-scanning model, wherein the cable is very clear and accurate; and the target object is in real time.

And S105, judging whether to perform safety early warning on the target object according to the real-time distance.

And judging whether to carry out safety early warning on the target object according to the obtained real-time distance, so that early warning can be timely carried out before safety problems occur in the transformer substation for people, vehicles, animals and other foreign objects.

In an optional implementation manner, fig. 2 is a flowchart for determining whether to perform a safety warning on a target object according to a real-time distance according to an embodiment of the present invention; as shown in fig. 2, the S105 includes:

s1051, identifying the object type of the target object;

in an optional embodiment, after the specific position of the target object in the substation is obtained, the category of the target object is identified; for example, whether a person, car, animal, etc. is identified.

S1052, acquiring a safety distance of the object type;

in an alternative embodiment, the safe distance of each type of target object is obtained, wherein the safe distance of each type of target object is different; in particular, the safety distances for different object types at different voltage levels are also different.

And S1053, comparing the real-time distance with the safety distance, and judging whether to perform safety early warning on the target object.

In an alternative embodiment, the real-time distance of each type of target object is compared with the safety distance thereof, and an additional 50% increase is given to each type of safety distance, resulting in the early warning distance. And sending an early warning alarm when the real-time distance of each type of target object reaches the early warning distance of the type, and sending a violation alarm when the real-time distance of each type of target object is smaller than the safety distance of the type.

Particularly, when the early warning is more, the real-time distance of each type of target object is arranged to the distance of the safety distance, and the alarm is started from the early warning with the nearest distance; and when the distances of various target objects are relatively close, alarming according to the requirement of the safe distance from large to small. The early warning mode comprises software platform information pushing and on-site acousto-optic warning. And the violation alarm also stores camera snapshot pictures and camera videos on the platform as a left certificate.

In an alternative implementation manner, fig. 3 is a flowchart of obtaining a rotation value and a translation value according to a pre-scan model and a calibration scan model according to an embodiment of the present invention, as shown in fig. 3, where S102 includes:

s1021, selecting a plurality of corresponding pairs of key points from the pre-scanning model and the correction scanning model respectively;

in an optional embodiment, for a certain position of the substation, the position is selected from the pre-scan model and the correction scan model respectively, and is marked as a key point Ppn and a key point Pcn under respective coordinates of the pre-scan model and the correction scan model; and selecting n key point pairs according to the model condition.

S1022, correcting the correction scanning model by matching each pair of key points to obtain the rotation value and the translation value.

In an alternative embodiment, each pair of the keypoints is matched, even if the corrected scan model is corrected through a certain rotation, translation and scaling, and the rotation value and the translation value are obtained.

In an optional embodiment, after S1022, the method includes:

s1023, extracting a certain area correction scanning model range in the pre-scanning model and an area scanning model within a preset distance nearby the certain area correction scanning model range; down-sampling the area scanning model and extracting the outer contour of the area scanning model after down-sampling;

in an optional embodiment, extracting a region scan model within a region correction scan model range and within 10m around the region correction scan model range in the pre-scan model to reduce subsequent computation, and performing down-sampling on the region scan model; specifically, the down-sampling is performed by using 10 × 10cm cubic grids, and the center of gravity of all points in each grid is calculated as a down-sampling result in the grid, so that the area scanning model after down-sampling is obtained. And extracting the outer contour of the area scanning model after down sampling for standby.

And S1024, performing quasi-correction scanning with second preset precision on the certain area again according to the down-sampled outer contour of the area scanning model and preset time, correcting the outer contour of the extracted quasi-correction scanning model, and updating the rotation value and the translation value.

The calibration scan and the identification scan of the second predetermined accuracy are both performed by using the same coarse scanning device. For the fine scanning device that scans of carrying out first predetermined precision, this rough scanning device stability is relatively poor. Such as the possibility of the rough-scanning device moving with external conditions, such as windy weather, self-weight effects, and human error touch; or because other field requirements may suggest altering the position of the rough-scanning device. Based on this, the rough scanning apparatus needs to be calibrated periodically. The calibration method comprises the following steps: the rotation value and the translation value are updated periodically.

Specifically, according to a preset time interval, the outer contour of the quasi-correction scanning model is extracted, the outer contour of the quasi-correction scanning model is corrected according to the down-sampled outer contour of the area scanning model, and a rotation value and a translation value are updated. In the present embodiment, the quasi-calibration scan is performed at 6 am. It should be noted that the time interval between two adjacent scans is the term of the correction scan, which is only an example of the present embodiment, and it is only for better describing the present embodiment, so the present invention is not limited thereto. Furthermore, when the grass in the transformer substation grows higher or when a device is placed in the transformer substation for one week temporarily, quasi-correction scanning is carried out on the transformer substation in real time to obtain a quasi-correction scanning model, and the quasi-correction scanning model is corrected to update the rotation value and the translation value.

In an optional embodiment, the S1024 includes:

s10241, performing quasi-correction scanning with second preset precision on the certain area again according to preset time, and extracting to obtain the outer contour of the quasi-correction scanning model;

s10242, judging whether the difference value between the outer contour of the quasi-correction scanning model and the outer contour of the correction scanning model exceeds a preset threshold value or not; if not, automatically correcting the quasi-correction scanning model, performing down-sampling on the outer contour of the corrected quasi-correction scanning model, and automatically correcting the outer contour of the area scanning model after down-sampling; and if so, manually correcting the quasi-correction scanning model.

In an alternative embodiment, the quasi-calibration scan model and the calibration scan model may not be the same, and some objects in the quasi-calibration scan model may have changed, such as some equipment cabinet doors being opened, a knife switch being opened to closed, etc. In some point cloud registration algorithms, it is considered that the two models to be registered should be identical, and the above-mentioned variations cause errors in the conventional algorithms. In order to eliminate the error and improve the running speed, only the outer contour of the quasi-correction scanning model is extracted during calculation, the outer contour of the quasi-correction scanning model is compared, and points with larger position difference are removed. When the remaining point cloud is less than 30% of the original profile of the corrected scan model, it is determined that automatic calibration cannot be performed, and manual calibration needs to be performed again.

Fig. 4 is a schematic structural diagram of a safety precaution device provided in an embodiment of the present invention, and as shown in fig. 4, the safety precaution device includes:

the pre-scanning unit 201 is configured to perform pre-scanning with a first preset precision on the substation to obtain a pre-scanning model;

the correction scanning unit 202 is configured to perform correction scanning with a second preset precision on the substation to obtain a correction scanning model; obtaining a rotation value and a translation value according to the pre-scanning model and the correction scanning model; wherein the accuracy of the pre-scan is higher than the accuracy of the correction scan;

in an optional embodiment, the first preset precision is 1mm, the second preset precision is 3mm, the transformer substation scene is finely scanned by using equipment with the scanning precision of 1mm to obtain a pre-scanning model, and the transformer substation scene is roughly scanned once by using equipment with the scanning precision of 3mm to obtain a correction scanning model; the cables in the pre-scanning model are very clear, and the correction scanning model is influenced by the scanning precision, so that the obtained cables have breakpoints and covering conditions;

at this time, the corrected scanning model and the pre-scanning model are corrected to obtain a rotation value and a translation value at the same time, and the obtained rotation value and translation value are used for subsequent calculation.

The correction identification unit 203 is used for performing subsequent multiple recognition scans with second preset precision on the transformer substation in real time, and correcting the recognition scan model obtained by each recognition scan according to the rotation value and the translation value to obtain a correction recognition model;

in an optional embodiment, the substation is subjected to subsequent multiple recognition scanning with second preset precision by using equipment with the scanning precision of 3 mm; for example, the subsequent multiple recognition scans are … … of 2 nd, 3 th, and 4 … … 99 th, and the recognition scan model obtained from each recognition scan is directly corrected according to the above rotation value and translation value to obtain a corrected recognition model, specifically, the point P (X, Y, Z) on the rough scan model is transformed according to the following formula to obtain the fine scan model P (X, Y, Z):

wherein (X, Y, Z) is a point on the rough scan model, (X, Y, Z) is a point on the fine scan model, R is the rotation value, and T is the translation value.

Furthermore, the interval time of every two adjacent identification scans is the same, and two identification scans can be realized within one second; assuming that there are 99 of the corrected recognition models; in the above 99 identification scans, there may be foreign objects such as people, cars, animals, etc., for example, a person appears in the 2 nd identification scan, a person and a car appear in the 20 th identification scan, a person and two cars appear in the 50 th identification scan, and a plurality of persons, a plurality of cars and a plurality of animals appear in the 99 th identification scan. In the process, each subsequent identification scanning is directly corrected according to the rotation value R and the translation value T obtained by the first correction, the rotation value and the translation value do not need to be recalculated, the correction time of operators is saved, and the correction speed is accelerated.

A calculating unit 204, configured to perform a difference operation on the calibration identification model and the pre-scanning model to obtain a target object; calculating the real-time distance from the target object to a cable in the pre-scanning model;

in an optional embodiment, the calibration identification model and the pre-scan model are subjected to a differential operation to remove a background in the substation, so as to obtain the target object, for example: humans, cars, animals, etc.; calculating the real-time distance from the target object to a cable in the pre-scanning model, wherein the cable is very clear and accurate; and the target object is in real time.

And the safety early warning unit 205 is configured to determine whether to perform safety early warning on the target object according to the real-time distance.

And judging whether to carry out safety early warning on the target object according to the obtained real-time distance, so that early warning can be timely carried out before safety problems occur in the transformer substation for people, vehicles, animals and other foreign objects.

In an optional implementation manner, fig. 5 is a schematic structural diagram of a safety precaution unit provided in an embodiment of the present invention, and as shown in fig. 5, the safety precaution unit 205 includes:

an identification subunit 2051 configured to identify an object class of the target object;

in an optional embodiment, after the specific position of the target object in the substation is obtained, the category of the target object is identified; for example, whether a person, car, animal, etc. is identified.

An obtaining subunit 2052, configured to obtain a safe distance of the object class;

in an alternative embodiment, the safe distance of each type of target object is obtained, wherein the safe distance of each type of target object is different; in particular, the safety distances for different object types at different voltage levels are also different.

And a safety pre-warning subunit 2053, configured to compare the real-time distance with the safety distance, and determine whether to perform safety pre-warning on the target object.

In an alternative embodiment, the real-time distance of each type of target object is compared with the safety distance thereof, and an additional 50% increase is given to each type of safety distance, resulting in the early warning distance. And sending an early warning alarm when the real-time distance of each type of target object reaches the early warning distance of the type, and sending a violation alarm when the real-time distance of each type of target object is smaller than the safety distance of the type.

Particularly, when the early warning is more, the real-time distance of each type of target object is arranged to the distance of the safety distance, and the alarm is started from the early warning with the nearest distance; and when the distances of various target objects are relatively close, alarming according to the requirement of the safe distance from large to small. The early warning mode comprises software platform information pushing and on-site acousto-optic warning. And the violation alarm also stores camera snapshot pictures and camera videos on the platform as a left certificate.

In an alternative implementation manner, fig. 6 is a schematic structural diagram of a calibration scanning unit according to an embodiment of the present invention, and as shown in fig. 6, the calibration scanning unit 202 includes:

a selecting subunit 2021, configured to select multiple corresponding pairs of key points from the pre-scanning model and the calibration scanning model respectively;

in an optional embodiment, for a certain position of the substation, the position is selected from the pre-scan model and the correction scan model respectively, and is marked as a key point Ppn and a key point Pcn under respective coordinates of the pre-scan model and the correction scan model; and selecting n key point pairs according to the model condition.

A matching subunit 2022, configured to correct the corrected scan model by matching each pair of the keypoints and obtain the rotation value and the translation value.

In an alternative embodiment, each pair of the keypoints is matched, even if the corrected scan model is corrected through a certain rotation, translation and scaling, and the rotation value and the translation value are obtained.

In an optional embodiment, the calibration scanning unit 202 further includes:

an extracting subunit 2023, configured to extract a region scan model within a region correction scan model range and a preset distance around the region correction scan model range in the pre-scan model; down-sampling the area scanning model and extracting the outer contour of the area scanning model after down-sampling;

in an optional embodiment, extracting a region scan model within a region correction scan model range and within 10m around the region correction scan model range in the pre-scan model to reduce subsequent computation, and performing down-sampling on the region scan model; specifically, the down-sampling is performed by using 10 × 10cm cubic grids, and the center of gravity of all points in each grid is calculated as a down-sampling result in the grid, so that the area scanning model after down-sampling is obtained. And extracting the outer contour of the area scanning model after down sampling for standby.

A corrector subunit 2024, configured to perform, according to the down-sampled outer contour of the region scan model, quasi-calibration scan with a second preset precision on the certain region again according to a preset time, correct the extracted outer contour of the quasi-calibration scan model, and update the rotation value and the translation value.

The calibration scan and the identification scan of the second predetermined accuracy are both performed by using the same coarse scanning device. For the fine scanning device that scans of carrying out first predetermined precision, this rough scanning device stability is relatively poor. Such as the possibility of the rough-scanning device moving with external conditions, such as windy weather, self-weight effects, and human error touch; or because other field requirements may suggest altering the position of the rough-scanning device. Based on this, the rough scanning apparatus needs to be calibrated periodically. The calibration method comprises the following steps: the rotation value and the translation value are updated periodically.

Specifically, according to a preset time interval, the outer contour of the quasi-correction scanning model is extracted, the outer contour of the quasi-correction scanning model is corrected according to the down-sampled outer contour of the area scanning model, and a rotation value and a translation value are updated. In the present embodiment, the quasi-calibration scan is performed at 6 am. It should be noted that the time interval between two adjacent scans is the term of the correction scan, which is only an example of the present embodiment, and it is only for better describing the present embodiment, so the present invention is not limited thereto. Furthermore, when the grass in the transformer substation grows higher or when a device is placed in the transformer substation for one week temporarily, quasi-correction scanning is carried out on the transformer substation in real time to obtain a quasi-correction scanning model, and the quasi-correction scanning model is corrected to update the rotation value and the translation value.

In an alternative embodiment, the syndrome unit 2024 includes:

performing quasi-correction scanning with second preset precision on the certain area again according to preset time, and extracting to obtain the outer contour of the quasi-correction scanning model;

judging whether the difference value between the outer contour of the quasi-correction scanning model and the outer contour of the correction scanning model exceeds a preset threshold value or not; if not, automatically correcting the quasi-correction scanning model, performing down-sampling on the outer contour of the corrected quasi-correction scanning model, and automatically correcting the outer contour of the area scanning model after down-sampling; and if so, manually correcting the quasi-correction scanning model.

In an alternative embodiment, the quasi-calibration scan model and the calibration scan model may not be the same, and some objects in the quasi-calibration scan model may have changed, such as some equipment cabinet doors being opened, a knife switch being opened to closed, etc. In some point cloud registration algorithms, it is considered that the two models to be registered should be identical, and the above-mentioned variations cause errors in the conventional algorithms. In order to eliminate the error and improve the running speed, only the outer contour of the quasi-correction scanning model is extracted during calculation, the outer contour of the quasi-correction scanning model is compared, and points with larger position difference are removed. When the remaining point cloud is less than 30% of the original profile of the corrected scan model, it is determined that automatic calibration cannot be performed, and manual calibration needs to be performed again.

The invention also provides a safety early warning system which comprises the safety early warning device.

The present invention also provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the above-described security pre-warning method.

The storage medium stores the software, and the storage medium includes but is not limited to: optical disks, floppy disks, hard disks, erasable memory, etc.

The invention has the beneficial effects that:

the invention provides a safety early warning method, which comprises the steps of scanning a transformer substation with a first preset precision and a second preset precision to obtain a pre-scanning model, a correction scanning model and a correction identification model, and judging the positions of human, vehicle, animal and other foreign objects in the transformer substation to ensure that the foreign objects can be early warned in time before safety problems occur, so that the problems of accuracy and real time are simultaneously ensured; and obtaining a rotation value and a translation value according to the pre-scanning model and the correction scanning model, performing subsequent recognition scanning with second preset precision on the transformer substation for multiple times, and directly correcting the recognition scanning model obtained by each recognition scanning according to the rotation value and the translation value to obtain a correction recognition model.

Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.