A camera-based Baoguang real-time AI automatic monitoring and recognition method

Abstract

This invention relates to a real-time AI-based automatic monitoring and identification method for aurora borealis based on a camera, comprising the following steps: Step A, selecting a camera and adjusting its rotation direction according to the changes in the solar azimuth and altitude angles at the observation point's geographical location, so that the camera's detection direction is parallel to the direct sunlight; Step B, the camera takes pictures at regular intervals to form photographs and saves them; Step C, storing multiple aurora photographs in the control system's database and using AI recognition technology for deep learning; Step D, defining the identification area of ​​the photographs based on color and morphological features, comparing the identification area with the aurora photographs in the database, and determining whether the identification area has a certain proportion of reliable aurora features, otherwise determining whether there is aurora. This invention can capture and identify aurora in real time, with good monitoring results, and can provide timely information and aurora landscape resource analysis for local tourism.

Description

Technical Field

[0001] The invention relates to a camera-based Baoguang real-time AI automatic monitoring and identification method. Background Art

[0002] The meteorological landscape of Baoguang has become an important resource for mountain tourism due to its special formation conditions. For example, Mount Emei attracts tourists with its frequent Baoguang. Baoguang, also known as Emei Baoguang, English name GLORY or Brocken Bow, is a peculiar atmospheric light phenomenon. When a person faces away from the sun, a colorful halo centered on his or her own shadow can be seen on the background of clouds and fog composed of small water droplets in front of the human body. This colorful halo is called Baoguang. The human eye or camera in the mountain can only observe Baoguang when the three points of "sun-person or camera-Baoguang" are in a straight line.

[0003] Since Baoguang has very strict requirements on the observation location, cloud and sunlight conditions, it is often fleeting and difficult to observe. Therefore, it is inevitable to miss it if we only rely on accidental photos taken by tourists or manual observations at high mountain meteorological stations, which cannot truly reflect the endowment of Baoguang mountain tourism resources in a certain area.

[0004] In view of this, the inventor of this case conducted in-depth research on the above-mentioned issues, which led to the emergence of this case. Summary of the invention

[0005] The object of the present invention is to provide an automatic monitoring and identification method capable of more effectively monitoring precious light in real time.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A camera-based Baoguang real-time AI automatic monitoring and recognition method comprises the following steps:

[0008] Step A, selecting a camera that can rotate in the horizontal direction and the vertical direction and can be controlled by a software controller, and synchronously adjusting the rotation direction of the camera according to the changes in the solar azimuth and altitude angle of the geographical location of the observation point, so that the detection direction of the camera is parallel to the direct sunlight;

[0009] Step B, the control system controls the camera to take pictures at regular intervals to form pictures, and save the pictures;

[0010] Step C, storing a plurality of Baoguang photos with Baoguang meteorological landscapes in a database of the control system, and using AI recognition technology for deep learning;

[0011] Step D, the control system frames the identification area of ​​the photograph according to the color and morphological characteristics, and compares the identification area with the Baoguang photos in the database. When the identification area has a certain proportion of credible Baoguang features, it is judged to be Baoguang, otherwise, it is judged to be without Baoguang.

[0012] As a preferred embodiment of the present invention, the control system assigns the solar azimuth and altitude angle at that moment to the camera every 10 seconds and takes pictures.

[0013] As a preferred embodiment of the present invention, the photo shooting time is from 5:00 to 19:00 every day.

[0014] As a preferred embodiment of the present invention, the rotation direction of the camera is controlled by the solar altitude angle and azimuth angle corresponding to the ephemeris stored in the control system. As a preferred embodiment of the present invention, when it is determined to be a precious light, the control system reminds through email notification or voice broadcast.

[0015] As a preferred embodiment of the present invention, after determining that it is a precious light, the control system starts the recording function, and the recording time is three minutes. If the precious light appears again within three minutes, the recording time is automatically extended, otherwise the recording is stopped.

[0016] As a preferred embodiment of the present invention, the certain percentage of credibility is set to 99%.

[0017] After adopting the technical solution of the present invention, the camera can rotate synchronously with the sun, capture the meteorological landscape and compare it with the photos in the database using AI recognition technology to determine whether the captured photo has Baoguang. After the present invention finds Baoguang, it can further prompt and record. The present invention can capture in real time according to the characteristics of Baoguang meteorological landscape, better monitor Baoguang, and provide timely information and Baoguang landscape resource analysis for local tourism. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram of the present invention.

[0019] In the figure:

[0020] Mountain 10 Sun 20

[0021] Mounting bracket 30 Camera 40

[0022] Cloud curtain area 50 Baoguang 60

[0023] Control system 70 Sunlight 80 DETAILED DESCRIPTION

[0024] In order to further explain the technical solution of the present invention, it is described in detail below in conjunction with embodiments.

[0025] Reference Figure 1 , a Baoguang real-time AI automatic monitoring and recognition method based on a camera, comprising the following steps:

[0026] Step A, select a camera 40 that can rotate in the horizontal and vertical directions and can be controlled by a software controller, and adjust the rotation direction of the camera 40 according to the changes in the azimuth and altitude angle of the sun 20 at the geographical location of the observation point, so that the detection direction of the camera 40 is parallel to the direct light from the sun 20. Relative to the sun 20, the earth has two main movements, one is that it revolves around the sun 20, and the tilt of the earth's axis causes the north-south angle (declination δ) of the sun 20 directly hitting the earth every day to change; the other is that the earth itself is also constantly rotating, thus forming the change of the sun 20 hour angle (i.e., azimuth angle ω) in a day, that is, the sun 20 rises in the east and sets in the west. These two movements cause the sun 20 rays directly hitting the earth to change all the time. For observers in the northern hemisphere, the sun 20 rises from the east and moves clockwise to the west and falls on the horizon, that is, the azimuth of the sun 20 changes. At the same time, the angle of the sun 20 rays directly hitting the earth, that is, the altitude of the sun 20, also experiences a process from 0° (the sun 20 rises in the east in the morning) → increases → maximum (noon) → decreases → 0° (the sun 20 sets in the evening). By setting the horizontal angle and the up and down angles of the camera 40 to the azimuth and altitude of the sun 20, respectively, the synchronous rotation of the camera 40 and the sun 20 is achieved.

[0027] Step B, the control system 70 controls the camera 40 to take pictures at regular intervals, and save the pictures; as a preferred embodiment of the present invention, the control system 70 assigns the azimuth and altitude of the sun 20 at that moment to the camera 40 every ten seconds and takes pictures, and the picture taking time is from 5:00 to 19:00 every day. The camera 40 working time is controlled and set every ten seconds from 5:00 to 19:00 every day to avoid ineffective work caused by the absence of the sun 20 at night; at the same time, since each picture is about 500K (depending on the resolution), in order to solve the capacity and calculation pressure caused by frequent shooting, and because the precious light usually lasts for several seconds, the camera 40 is set to rotate once every ten seconds and take a picture and save it. The ten-second working time interval can meet the real-time requirements of effective tracking and capturing.

[0028] Step C, storing multiple Baoguang photos with Baoguang meteorological landscapes in the database of the control system 70, and using AI recognition technology for deep learning; the control system 70 continuously trains and memorizes the photos in the database. The more data there is, the more Baoguang information can be memorized by the control system 70, and the higher the recognition rate.

[0029] Step D, the control system 70 frames the identification area of ​​the photographed photo according to the color and shape characteristics, and compares the identification area with the Baoguang photos in the database. When the identification area has a certain proportion (99% in the embodiment) of the Baoguang features with a certain degree of credibility, it is judged as Baoguang, otherwise, it is judged as no Baoguang.

[0030] As a preferred embodiment of the present invention, the rotation direction of the camera 40 is controlled by the altitude angle and azimuth angle of the sun 20 corresponding to the ephemeris stored in the control system 70. The present invention can also calculate the altitude angle and azimuth angle of the sun 20 by the longitude, latitude and time point of the shooting location, which is common knowledge in the field of astronomy.

[0031] As a preferred embodiment of the present invention, after judging that it is a treasure light, the control system 70 reminds through email notification or voice broadcast. As a preferred embodiment of the present invention, after judging that it is a treasure light, the control system 70 starts the recording function, and the recording time is three minutes. If the treasure light appears again within three minutes, the recording time is automatically extended, otherwise the recording is stopped.

[0032] Figure 1 This is a schematic diagram of the present invention. For ease of viewing, the components or reference objects are not shown in actual proportion. 10 represents the mountain, 20 represents the sun, 50 represents the cloud curtain area, 60 represents the precious light, and 80 represents the sun's rays. A mounting frame 30 is arranged on the mountain 10, and the camera 30 is mounted on the mounting frame 30. The camera of the present invention can be directly purchased from the market, for example, a Hikvision dome camera, model DS-2DC4423IW-D.

[0033] The camera-based Baoguang real-time AI automatic monitoring and identification method of this application has been used in Jiuxian Mountain in Dehua, Fujian, and achieved good monitoring results.

[0034] The product form of the present invention is not limited to the embodiments of this case. Any appropriate changes or modifications made by anyone with similar ideas should be deemed to be within the patent scope of the present invention.

Claims

1. A camera-based Baoguang real-time AI automatic monitoring and recognition method, characterized in that : Includes the following steps: Step A, select a camera that can rotate in the horizontal direction and the vertical direction and is controlled by the software controller, and adjust the rotation direction of the camera according to the changes in the solar azimuth and altitude angle of the geographical location of the observation point so that the detection direction of the camera is parallel to the direct sunlight; Step B, the control system controls the camera to take pictures at regular intervals to form pictures, and save the pictures; Step C, storing a plurality of Baoguang photos with Baoguang meteorological landscapes in a database of the control system, and using AI recognition technology for deep learning; Step D, the control system frames the identification area of ​​the photograph according to the color and morphological characteristics, and compares the identification area with the Baoguang photos in the database. When the identification area has a certain proportion of credible Baoguang features, it is judged to be Baoguang, otherwise, it is judged to be without Baoguang.

2. A camera-based Baoguang real-time AI automatic monitoring and identification method as claimed in claim 1, characterized in that :The control system assigns the solar azimuth and altitude angle at that moment to the camera every 10 seconds and takes pictures.

3. A camera-based Baoguang real-time AI automatic monitoring and identification method as claimed in claim 2, characterized in that :Photo shooting time is from 5:00 to 19:00 every day.

4. A camera-based Baoguang real-time AI automatic monitoring and identification method as claimed in claim 3, characterized in that : The rotation direction of the camera is controlled by the solar altitude angle and azimuth angle corresponding to the ephemeris stored in the control system.

5. A camera-based Baoguang real-time AI automatic monitoring and identification method as claimed in claim 4, characterized in that : After judging it as Baoguang, the control system will remind you through email notification or voice broadcast.

6. A camera-based Baoguang real-time AI automatic monitoring and identification method as claimed in claim 5, characterized in that : After determining that it is a treasure light, the control system starts the recording function, and the recording time is three minutes. If the treasure light appears again within three minutes, the recording time will be automatically extended, otherwise the recording will be stopped.

7. The camera-based Baoguang real-time AI automatic monitoring and identification method as claimed in claim 1, characterized in that :The proportion of the certain proportion of credibility is 99%.

Images