Building lighting optimization and energy-saving control system based on artificial intelligence

Abstract

The present invention relates to the field of lighting energy-saving control technology, and specifically to an artificial intelligence-based lighting optimization and energy-saving control system for buildings, comprising: an image acquisition module for acquiring indoor images; an area screening module for obtaining shadow connected domains and non-dynamic areas based on the number of pixels in the overlapping area of ​​suspected shadow connected domains in the same location area in indoor images at each moment and at several previous moments; a lighting change analysis module for obtaining a light intensity adjustment factor based on the difference between the overall contrast of the same non-dynamic area in indoor images at all moments and the slope and direction of the grayscale fitting line of all pixels in the same non-dynamic area; and an optimization adjustment module for adjusting the original light intensity using the light intensity adjustment factor to obtain the adjusted light intensity. The present invention reduces the impact of hysteresis during lighting adjustment and reduces energy consumption.

Description

Technical Field

[0001] The present invention relates to the technical field of lighting energy-saving control, and in particular to an artificial intelligence-based lighting optimization and energy-saving control system for buildings. Background Art

[0002] In buildings, light regulation has a significant impact on user comfort, eye health, energy management, building function optimization, and sustainable development. Existing building lighting optimization systems usually rely on preset times and thresholds to control the on and off of lighting systems, but this method cannot dynamically adapt to indoor lighting needs according to different weather or seasonal conditions. Most existing building lighting optimization systems use preset times and thresholds to control the lighting system to turn on or off, which cannot adapt to lighting needs in different weather or seasons. Although some systems can adjust the brightness of indoor lighting according to the intensity of outdoor natural light, these systems lack analysis of lighting change trends and cannot predict lighting changes in advance, resulting in a lag in the adjustment of indoor light brightness, affecting the activity comfort and work efficiency of indoor occupants.

[0003] Existing lighting control systems, after detecting changes in outdoor natural light intensity, are often unable to instantly adjust indoor lighting levels, resulting in unstable indoor lighting levels. This delayed response impacts indoor comfort, especially when lighting needs fluctuate rapidly. The inability to provide appropriate indoor lighting in a timely manner reduces system efficiency and increases energy consumption. Summary of the Invention

[0004] The present invention provides an artificial intelligence-based building lighting optimization and energy-saving control system to solve the problem of hysteresis in lighting adjustment of existing energy-saving control systems.

[0005] The artificial intelligence-based building lighting optimization and energy-saving control system of the present invention adopts the following technical solutions:

[0006] One embodiment of the present invention provides an artificial intelligence-based building lighting optimization and energy-saving control system, which includes the following modules:

[0007] Image acquisition module: used to obtain indoor images at several consecutive moments;

[0008] Region screening module: used to evenly divide the indoor image at each moment to obtain several position areas; perform edge detection on each position area in the indoor image at each moment to obtain the suspected shadow connected domain of each position area in the indoor image at each moment; screen out the shadow connected domain from the suspected shadow connected domain based on the number of pixels in the overlapping area of ​​the suspected shadow connected domain in the same position area in the indoor image at each moment and the previous several moments; screen out the non-dynamic area from all position areas of the indoor image based on the number of edge pixels in the suspected shadow connected domain of each position area in the indoor image at each moment and the number of edge pixels in the overlapping area of ​​all shadow connected domains in the indoor image at each moment and the previous several moments, respectively;

[0009] Lighting change analysis module: This module is used to obtain the overall contrast of each non-dynamic area in the indoor image at each moment based on the grayscale histogram of each non-dynamic area in the indoor image at each moment; obtain the grayscale fitting line for each pixel; obtain the degree of change in the natural light intensity in the building based on the difference in the overall contrast of the same non-dynamic area in the indoor image at all moments and the slope and direction of the grayscale fitting line for all pixels in the same non-dynamic area; and perform negative mapping on the degree of change in the natural light intensity in the building to obtain the light intensity adjustment factor;

[0010] Optimization and adjustment module: used to obtain the original light intensity, adjust the original light intensity through the light intensity adjustment factor, and obtain the adjusted light intensity.

[0011] Furthermore, the uniformly dividing the indoor image at each moment to obtain a plurality of position regions; performing edge detection on each position region in the indoor image at each moment to obtain a suspected shadow connected domain of each position region in the indoor image at each moment includes:

[0012] Divide the indoor image at each moment equally into Then the indoor image at each moment is obtained location areas; among them, Indicates the preset area quantity parameter;

[0013] The canny edge detection algorithm is used to perform edge detection on each position area in the indoor image at each moment, and all edge pixels in each position area in the indoor image at each moment are obtained; the closed area formed by the edge pixels in each position area in the indoor image at each moment is recorded as the suspected shadow connected domain of each position area in the indoor image at each moment.

[0014] Furthermore, the method of screening out shadow connected domains from suspected shadow connected domains based on the number of pixels in the overlapping area of ​​the suspected shadow connected domains at each moment and the same position area in the indoor images at several previous moments includes:

[0015]

[0016] Where, Indicates the In the indoor image at the moment The number of pixels in the suspected shadow connected domain of the location area, Indicates the The moment and In the indoor image at the moment The number of pixels in the overlapping area of ​​the suspected shadow connected domain in the location area, Indicates the In the indoor image at the moment The possibility that the location area is a suspected non-dynamic area, Indicates the number of moments before each moment;

[0017] The probability of the suspected non-dynamic area being greater than or equal to the preset first threshold Each position area in the indoor image at each moment is recorded as a suspected non-dynamic area; the suspected shadow connected domain in the suspected non-dynamic area is recorded as a shadow connected domain.

[0018] Furthermore, the non-dynamic area is screened out from all position areas of the indoor image based on the number of edge pixels in the suspected shadow connected domain of each position area in the indoor image at each moment, the number of edge pixels in the overlapping area of ​​all shadow connected domains in the indoor images at each moment and at several previous moments, and the proportion of all edge pixels in each position area in the indoor image at each moment to the total number of all edge pixels, including:

[0019]

[0020] Where, Indicates the In the indoor image at the moment The number of edge pixels in the suspected shadow connected domain of the location area, Indicates the The indoor image at the moment and all reference moments The number of edge pixels in the overlapping area of ​​all shadow connected domains in the location area, Indicates the In the indoor image at the moment The total number of all edge pixels within the location area; Represents the total number of all indoor images taken in time series; Indicates the indoor image The possibility that the location area is a non-dynamic area;

[0021] Among them, the moments, recorded as the reference moment of each moment; among them, Indicates the preset reference quantity;

[0022] The probability of the non-dynamic area being greater than or equal to the preset first threshold The location area in the indoor image at each moment is recorded as a non-dynamic area.

[0023] Furthermore, the step of obtaining the overall contrast of each non-dynamic area in the indoor image at each moment according to the grayscale histogram of each non-dynamic area in the indoor image at each moment; and obtaining the grayscale fitting line of each pixel point includes:

[0024] Obtaining a grayscale histogram of each non-dynamic area in the indoor image at each moment, calculating the standard deviation of the number of pixels corresponding to all grayscale values ​​of the grayscale histogram of each non-dynamic area in the indoor image at each moment, and recording the reciprocal of the standard deviation as the overall contrast of each non-dynamic area in the indoor image at each moment;

[0025] The grayscale values ​​of all pixels at the same position in the indoor image at all times are obtained, and the grayscale values ​​of all pixels at the same position are linearly fitted using the least squares method to obtain a grayscale fitting line for each pixel.

[0026] Furthermore, obtaining the degree of change in the natural light intensity in the building interior according to the difference between the overall contrasts of the same non-dynamic area in the indoor images at all times and the slope and direction of the grayscale fitting line of all pixels in the same non-dynamic area includes:

[0027] The degree of illumination change in each non-dynamic area in the indoor image is obtained based on the difference in overall contrast between the same non-dynamic area in the indoor image at all times and the slope of the grayscale fitting line of all pixels in the same non-dynamic area. The degree of change in natural light intensity in the building interior is obtained based on the degree of illumination change in each non-dynamic area in the indoor image and the slope direction of the grayscale fitting line of all pixels in the non-dynamic area.

[0028] Furthermore, obtaining the degree of illumination change of each non-dynamic area in the indoor image based on the difference between the overall contrasts of the same non-dynamic area in the indoor image at all times and the slope of the grayscale fitting line of all pixels in the same non-dynamic area includes:

[0029]

[0030] Where, Indicates the In the indoor image at the moment The overall contrast of the non-dynamic area, Indicates the In the indoor image at the moment The overall contrast of the non-dynamic area, represents the total number of all indoor images taken in time series, Indicates the The maximum value of the slope of the grayscale fitting line of all pixels in the non-dynamic area, represents the linear normalization function, Indicates the indoor image The degree of illumination change in a non-dynamic area.

[0031] Furthermore, obtaining the degree of change in natural light intensity in the building interior according to the degree of change in illumination in each non-dynamic area in the indoor image and the slope direction of the grayscale fitting line of all pixels in the non-dynamic area includes:

[0032]

[0033] Where, Indicates the indoor image The degree of illumination change in a non-dynamic area, Indicates the The maximum value of the slope of the grayscale fitting line of all pixels in the non-dynamic area, Indicates the absolute value symbol, represents the set of all non-dynamic regions, Indicates taking the minimum value from the set of all non-dynamic areas, Indicates the degree of change in natural light intensity in a building.

[0034] Furthermore, the negative mapping of the degree of change in the intensity of natural light in the building to obtain the light intensity adjustment factor includes:

[0035]

[0036] Where, Indicates the degree of change in the intensity of natural light in a building. Indicates the light intensity adjustment factor.

[0037] Furthermore, the adjusting the original light intensity by the light intensity adjustment factor to obtain the adjusted light intensity includes:

[0038]

[0039] Where, Indicates the original light intensity. Indicates adjusting the light intensity. Indicates the light intensity adjustment factor.

[0040] The beneficial effects of the technical solution of the present invention are: evenly dividing the indoor image at each moment to obtain several position areas; performing edge detection on each position area in the indoor image at each moment to obtain a suspected shadow connected domain for each position area in the indoor image at each moment, thereby improving the accuracy of interference analysis of local area movement factors; screening out shadow connected domains from suspected shadow connected domains based on the number of pixel points in the overlapping area of ​​the suspected shadow connected domain in the same position area in the indoor image at each moment and the previous several moments; screening out shadow connected domains from all position areas in the indoor image based on the number of edge pixel points in the suspected shadow connected domain in each position area in the indoor image at each moment, the number of edge pixel points in the overlapping area of ​​all shadow connected domains in the indoor image at each moment and the previous several moments, and the proportion of the edge pixel points in each position area in the indoor image at each moment to the total number of all edge pixel points. Non-dynamic areas are screened out, thereby improving the accuracy of non-dynamic area analysis; based on the grayscale histogram of each non-dynamic area in the indoor image at each moment, the overall contrast of each non-dynamic area in the indoor image at each moment is obtained, thereby improving the accuracy of local contrast analysis; the grayscale fitting line of each pixel point is obtained; based on the difference between the overall contrasts of the same non-dynamic area in the indoor images at all moments, the slope and direction of the grayscale fitting line of all pixels in the same non-dynamic area, the degree of change in the natural light intensity in the building interior is obtained; the degree of change in the natural light intensity in the building interior is negatively mapped to obtain the light intensity adjustment factor, and the direction and accuracy of light intensity optimization are improved through the light intensity adjustment factor; the original light intensity is adjusted through the light intensity adjustment factor to obtain the adjusted light intensity, thereby reducing the impact of lag during light adjustment and reducing energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0042] Figure 1 This is a module flow chart of the artificial intelligence-based building lighting optimization and energy-saving control system of the present invention. DETAILED DESCRIPTION

[0043] To further illustrate the technical means and effectiveness of the present invention in achieving its intended objectives, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effectiveness of the artificial intelligence-based building lighting optimization and energy-saving control system proposed by the present invention. In the following description, different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0044] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0045] The specific scheme of the building lighting optimization and energy-saving control system based on artificial intelligence provided by the present invention is described in detail below with reference to the accompanying drawings.

[0046] See also Figure 1 , which shows a module flow chart of an artificial intelligence-based building lighting optimization and energy-saving control system provided by one embodiment of the present invention. The system includes the following modules:

[0047] Image acquisition module 101.

[0048] It's important to note that cameras installed near each light capture time-series images of the interior at regular intervals to analyze lighting trends. By comparing light intensity in consecutive images and analyzing the time series, the system predicts natural light trends and adjusts artificial lighting brightness accordingly. The system provides real-time feedback and optimizes lighting control strategies to ensure indoor lighting remains within the optimal range, while improving energy efficiency and comfort.

[0049] Specifically, a camera is installed near a light in the room, and then the camera is To collect images, a set of indoor lighting images in time sequence is obtained. In each moment, an indoor lighting image is collected. In this embodiment, the time interval is preset. Minutes, in this embodiment, the preset time interval There is no specific limitation and implementers can decide based on specific circumstances.

[0050] The indoor lighting image is grayscaled to obtain an indoor image. The process of grayscaled indoor lighting image is a well-known technology and will not be described in detail here.

[0051] At this point, indoor images at several consecutive moments are obtained.

[0052] Region screening module 102.

[0053] It should be noted that when collecting images, if there is no movement interference indoors in the building, the trend of lighting changes can be analyzed through time-series indoor images to make adaptive adjustments to the lighting in the building in advance; however, in reality, there may be movement of people or objects. At this time, there will be deviations in the lighting trend analysis directly through time-series images. Therefore, the non-dynamic areas in the indoor images are first screened, and the lighting trend analysis is performed on the non-dynamic areas. This can improve the accuracy of the lighting trend change analysis.

[0054] It should be further explained that, because indoor human activities are usually concentrated in certain locations, while other locations change relatively seldom, for example, in an office building, activities at desks are more frequent, while the locations of some large equipment and instruments usually do not change much, so it is necessary to first divide the indoor image into blocks, and filter out the non-dynamic areas in the indoor image by the frequency of human activities in each area; that is, the more frequent the human activities in an area, the greater the difference in grayscale changes in the area over time; and when the human activities in an area are infrequent or even there is no activity, the smaller the difference in grayscale changes in the area over time.

[0055] Specifically, the indoor image at each moment is divided into equal parts, and the indoor image at each moment is obtained location areas; among them, Indicates the preset area quantity parameter; wherein, in this embodiment, the preset area quantity parameter , in this embodiment, the preset area number parameter There is no specific limitation and implementers can decide based on specific circumstances.

[0056] It should be noted that, because the change of natural lighting usually does not change significantly in a short period of time, the non-dynamic area in the indoor image can be judged and analyzed by the grayscale distribution difference or edge overlap of the same position area of ​​the indoor image at multiple consecutive moments in time sequence; in order to determine the edge overlap of the same position area of ​​adjacent indoor images in time sequence, edge detection is first performed on the indoor image to obtain the edge pixel points of the indoor image at each moment, and the edge overlap is determined and analyzed by the distribution of the edge pixel points.

[0057] Specifically, we use the Canny edge detection algorithm to perform edge detection on each location region in the indoor image at each moment, obtaining all edge pixels within each location region. The closed region formed by these edge pixels within each location region in the indoor image at each moment is recorded as the suspected shadow connected domain for each location region in the indoor image at each moment. This method now obtains the suspected shadow connected domain for each location region in the indoor image at each moment.

[0058] The canny edge detection algorithm is a well-known technology and will not be described in detail here.

[0059] It should be noted that whether the position area is a suspected non-dynamic area is determined by analyzing the degree of overlap of the suspected shadow connected domains corresponding to the same position area in the indoor images at several consecutive moments in time; that is, the greater the degree of overlap of the suspected shadow connected domains corresponding to the same position area, the greater the possibility that the position area is a suspected non-dynamic area; the smaller the degree of overlap of the suspected shadow connected domains corresponding to the same position area, the smaller the possibility that the position area is a suspected non-dynamic area.

[0060] Specifically, the probability that each location area in the indoor image at each moment is a suspected non-dynamic area is obtained based on the number of pixels in the overlapping area of ​​the suspected shadow connected domain at each moment and the same location area in the indoor image at several previous moments. The probability of the suspected non-dynamic area is specifically expressed by the formula:

[0061]

[0062] Where, Indicates the In the indoor image at the moment The number of pixels in the suspected shadow connected domain of the location area, Indicates the The moment and In the indoor image at the moment The number of pixels in the overlapping area of ​​the suspected shadow connected domain in the location area, Indicates the In the indoor image at the moment The possibility that the location area is a suspected non-dynamic area; It represents the number of moments before each moment, wherein in this embodiment, the number of moments before each moment is 10. In this embodiment, the number of moments before each moment is not specifically limited, and the implementer may determine it according to the specific situation.

[0063] in, It represents the number of pixels in the overlapping area of ​​the suspected shadow connected domain in the same position area in the indoor image at each moment and the previous moment, and its proportion in the total number of pixels in the suspected shadow connected domain in the same position area in the indoor image at each moment. The larger the proportion, the greater the degree of overlap; the smaller the proportion, the smaller the degree of overlap, which means that there is a person walking in the position area.

[0064] The probability of the suspected non-dynamic area being greater than or equal to the preset first threshold Each position area in the indoor image at each moment is recorded as a suspected non-dynamic area; the suspected shadow connected domain in the suspected non-dynamic area is recorded as a shadow connected domain. , wherein in this embodiment, the first threshold is preset There is no specific limitation and implementers can decide based on specific circumstances.

[0065] It should be noted that the greater the degree of overlap of shadow areas of static objects at different times in the same position area, the greater the possibility that the position area is a non-dynamic area. Conversely, the less likely the position area is a non-dynamic area.

[0066] Specifically, the probability that each position area in the indoor image is a non-dynamic area is obtained based on the number of edge pixels in the suspected shadow connected domain of each position area in the indoor image at each moment, the number of edge pixels in the overlapping area of ​​all shadow connected domains in the indoor images at each moment and all reference moments, and the proportion of all edge pixels in each position area in the indoor image at each moment to the total number of all edge pixels. The probability of the non-dynamic area is specifically expressed by the formula:

[0067]

[0068] Where, Indicates the In the indoor image at the moment The number of edge pixels in the suspected shadow connected domain of the location area, Indicates the The indoor image at the moment and all reference moments The number of edge pixels in the overlapping area of ​​all shadow connected domains in the location area, Indicates the In the indoor image at the moment The total number of all edge pixels within the location area; Represents the total number of all indoor images taken in time series; Indicates the indoor image The possibility that the location area is a non-dynamic area. moments, recorded as the reference moment of each moment; among them, Indicates a preset reference quantity; wherein, in this embodiment, the preset reference quantity , in this embodiment, the preset reference quantity There is no specific limitation and implementers can decide based on specific circumstances.

[0069] in, Indicates the ratio of edge pixels in the suspected shadow connected domain to all edge pixels in the same location area. The larger the ratio, the greater the possibility that the location area is a non-dynamic area. Conversely, the smaller the ratio, the smaller the possibility that the location area is a non-dynamic area. It indicates the ratio of edge pixels corresponding to stationary objects to all edge pixels in the same location area. The larger the ratio, the greater the possibility that the location area is a non-dynamic area. Conversely, the smaller the ratio, the smaller the possibility that the location area is a non-dynamic area.

[0070] The probability of the non-dynamic area being greater than or equal to the preset first threshold The location area in the indoor image at each moment is recorded as a non-dynamic area.

[0071] So far, the non-dynamic area in the indoor image is obtained through the above method.

[0072] Lighting change analysis module 103.

[0073] It should be noted that because the changes in natural light intensity are usually gradual and will not change suddenly, the changing trend of natural light intensity in the current period can be obtained based on the changes in light intensity in multiple indoor images taken continuously, thereby predicting the degree of change in natural light intensity.

[0074] It should be further explained that when the indoor brightness is strong, the colors of the captured images will be more vivid, that is, the image contrast will be higher. Therefore, the changes in the grayscale distribution in the continuous image can be used to obtain the changes in the light intensity in the area. Because the captured image may have relatively simple colors, such as floor tiles, white walls, etc., when the light intensity changes, the grayscale changes of the pixels in the image are consistent. Analyzing only based on the grayscale distribution will mistakenly identify such situations as no change in light. Therefore, it is also necessary to combine the grayscale changes of individual pixels to calculate the accurate changes in light intensity. Because when the light is dim, some objects that were originally lighter in color will also become darker, that is, the grayscale value is lower. When the light increases, these objects will gradually show their original color, that is, the grayscale value will gradually increase. Therefore, based on the image contrast intensity, the light intensity change trend is calculated by combining the grayscale changes of individual pixels in the image.

[0075] Specifically, a grayscale histogram is obtained for each non-dynamic area in the indoor image at each moment. The standard deviation of the number of pixels corresponding to all grayscale values ​​in the grayscale histogram of each non-dynamic area in the indoor image at each moment is calculated. The inverse of the standard deviation is recorded as the overall contrast of each non-dynamic area in the indoor image at each moment. The process of obtaining the grayscale histogram is well known and will not be described in detail here.

[0076] It's important to note that the grayscale value variation of each pixel in an image is not only related to light intensity but also affected by the object's color. For darker objects, their grayscale value fluctuates less with changing lighting. Meanwhile, lighter objects may not accurately represent their true color in weaker lighting, appearing as darker grayscale values. As lighting increases, the grayscale value fluctuates more significantly. Therefore, by analyzing the grayscale variation of each pixel across multiple captured images, we can more accurately determine changes in regional lighting, especially for lighter objects, where sensitivity to lighting changes is higher.

[0077] It's also worth noting that grayscale changes in light-colored objects are more likely to reflect changing trends in natural lighting, as their brightness is more sensitive to changes in lighting. Dark objects, on the other hand, reflect less light, so their grayscale changes are less responsive to lighting. Therefore, if the goal is to infer changes in light intensity from grayscale changes in an image, light-colored objects are more representative.

[0078] Specifically, the grayscale values ​​of all pixels at the same position in the indoor image at all times are obtained, and the grayscale values ​​of all pixels at the same position are linearly fitted using the least squares method to obtain a grayscale fitting line for each pixel; wherein the least squares method is a well-known technology and will not be described in detail here.

[0079] The degree of illumination change in each non-dynamic area in the indoor image is obtained based on the difference in overall contrast between the same non-dynamic area in the indoor image at all times and the slope of the grayscale fitting line of all pixels in the same non-dynamic area. The degree of illumination change is specifically expressed by the formula:

[0080]

[0081] Where, Indicates the In the indoor image at the moment The overall contrast of the non-dynamic area, Indicates the In the indoor image at the moment The overall contrast of the non-dynamic area, represents the total number of all indoor images taken in time series, Indicates the The maximum value of the slope of the grayscale fitting line of all pixels in the non-dynamic area, represents the linear normalization function, Indicates the indoor image The degree of illumination change in a non-dynamic area.

[0082] in, It represents the difference in the overall contrast of the same non-dynamic area in the indoor image at all adjacent moments in the time series. When the difference is larger, it means that the degree of illumination change in each non-dynamic area is greater, and vice versa, it means that the degree of illumination change in each non-dynamic area is smaller. When the maximum value of the slope of the grayscale fitting line of all pixels in each non-dynamic area is larger, it means that the degree of illumination change in each non-dynamic area is greater, and vice versa, it means that the degree of illumination change in each non-dynamic area is smaller.

[0083] It should be noted that, because the distances between different non-dynamic areas and windows in the captured images are different, in order to ensure that each non-dynamic area has sufficient lighting, the lighting adjustment should be based on the light intensity of the area with the lowest natural light intensity during the current period; because only by analyzing all other non-dynamic areas based on the lighting changes in the non-dynamic area with the lowest natural light intensity can it be ensured that all non-dynamic areas can be adaptively adjusted when the natural light changes.

[0084] It should be further explained that since the change of light can be in two directions, namely, enhancement and reduction, in order to ensure that there is no error in the adjustment process of the light, the direction of change of natural light is determined by the change direction of the slope of the grayscale fitting line of all pixels in the non-dynamic area.

[0085] Specifically, the degree of change in the natural light intensity in the building is obtained based on the degree of illumination change in each non-dynamic area in the indoor image and the slope direction of the grayscale fitting line of all pixels in the non-dynamic area. The degree of change in the natural light intensity in the building is specifically expressed by the formula:

[0086]

[0087] Where, Indicates the indoor image The degree of illumination change in a non-dynamic area, Indicates the The maximum value of the slope of the grayscale fitting line of all pixels in the non-dynamic area, Indicates the absolute value symbol, represents the set of all non-dynamic regions, Indicates taking the minimum value from the set of all non-dynamic areas, Indicates the degree of change in natural light intensity in a building.

[0088] in, It is used to define the direction of change of natural light, indicating that when natural light increases, the degree of change is positive, and when natural light decreases, the degree of change is negative.

[0089] It should be noted that when the natural light gradually decreases in time, the indoor lighting should be increased; when the natural light gradually increases in time, the indoor lighting should be reduced. Therefore, adjustment is made by negatively mapping the slope of the grayscale fitting line of the pixel points in the non-dynamic area.

[0090] Specifically, the degree of change in the natural light intensity in the building is negatively mapped to obtain a light intensity adjustment factor; the light intensity adjustment factor is specifically expressed by the formula:

[0091]

[0092] Where, Indicates the degree of change in the intensity of natural light in a building. Indicates the light intensity adjustment factor.

[0093] At this point, the light intensity adjustment factor is obtained through the above method.

[0094] Optimization and adjustment module 104.

[0095] The original light intensity is obtained, and the original light intensity is adjusted by the light intensity adjustment factor to obtain the adjusted light intensity; the adjusted light intensity is specifically expressed by the formula:

[0096]

[0097] Where, Indicates the original light intensity. Indicates adjusting the light intensity. Indicates the light intensity adjustment factor.

[0098] At this point, this embodiment is completed.

[0099] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An artificial intelligence-based building lighting optimization and energy-saving control system, characterized by: The system includes the following modules: Image acquisition module: used to obtain indoor images at several consecutive moments; Region screening module: used to evenly divide the indoor image at each moment to obtain several location regions; perform edge detection on each location region in the indoor image at each moment to obtain the suspected shadow connected domain of each location region in the indoor image at each moment; The non-dynamic area is obtained based on the number of edge pixels in the suspected shadow connected domain of each position area in the indoor image at each moment, the number of edge pixels in the overlapping area of ​​all shadow connected domains in the indoor images at each moment and several previous moments, and the proportion of all edge pixels in each position area in the indoor image at each moment to the total number of all edge pixels, including: Where, Indicates the In the indoor image at the moment The number of edge pixels in the suspected shadow connected domain of the location area, Indicates the The indoor image at the moment and all reference moments The number of edge pixels in the overlapping area of ​​all shadow connected domains in the location area, Indicates the In the indoor image at the moment The total number of all edge pixels within the location area; Represents the total number of all indoor images taken in time series; Indicates the indoor image The possibility that the location area is a non-dynamic area; The preset reference number of moments before each moment is recorded as the reference moment of each moment; The probability of the non-dynamic area being greater than or equal to the preset first threshold The location area in the indoor image at each moment is recorded as the non-dynamic area; Lighting change analysis module: This module is used to obtain the overall contrast of each non-dynamic area in the indoor image at each moment based on the grayscale histogram of each non-dynamic area in the indoor image at each moment; obtain the grayscale fitting line for each pixel; obtain the degree of change in the natural light intensity in the building based on the difference in the overall contrast of the same non-dynamic area in the indoor image at all moments and the slope and direction of the grayscale fitting line for all pixels in the same non-dynamic area; and perform negative mapping on the degree of change in the natural light intensity in the building to obtain the light intensity adjustment factor; Optimization and adjustment module: used to obtain the original light intensity, adjust the original light intensity through the light intensity adjustment factor, and obtain the adjusted light intensity.

2. The artificial intelligence-based building lighting optimization and energy-saving control system according to claim 1 is characterized in that: The indoor image at each moment is evenly divided to obtain a plurality of position areas; Perform edge detection on each location area in the indoor image at each moment to obtain the suspected shadow connected domain of each location area in the indoor image at each moment, including: Divide the indoor image at each moment equally into Then the indoor image at each moment is obtained location areas; among them, Indicates the preset area quantity parameter; The canny edge detection algorithm is used to perform edge detection on each position area in the indoor image at each moment, and all edge pixels in each position area in the indoor image at each moment are obtained; the closed area formed by the edge pixels in each position area in the indoor image at each moment is recorded as the suspected shadow connected domain of each position area in the indoor image at each moment.

3. The artificial intelligence-based building lighting optimization and energy-saving control system according to claim 1 is characterized in that: The method of screening out shadow connected domains from suspected shadow connected domains based on the number of pixels in the overlapping area of ​​the suspected shadow connected domains at each moment and the same position area in the indoor images at several previous moments includes: Where, Indicates the In the indoor image at the moment The number of pixels in the suspected shadow connected domain of the location area, Indicates the The moment and In the indoor image at the moment The number of pixels in the overlapping area of ​​the suspected shadow connected domain in the location area, Indicates the In the indoor image at the moment The possibility that the location area is a suspected non-dynamic area, Indicates the number of moments before each moment; The probability of the suspected non-dynamic area being greater than or equal to the preset first threshold Each position area in the indoor image at each moment is recorded as a suspected non-dynamic area; the suspected shadow connected domain in the suspected non-dynamic area is recorded as a shadow connected domain.

4. The artificial intelligence-based building lighting optimization and energy-saving control system according to claim 1 is characterized in that: The method further comprises obtaining the overall contrast of each non-dynamic area in the indoor image at each moment according to the grayscale histogram of each non-dynamic area in the indoor image at each moment; Get the grayscale fitting line for each pixel, including: Obtaining a grayscale histogram of each non-dynamic area in the indoor image at each moment, calculating the standard deviation of the number of pixels corresponding to all grayscale values ​​of the grayscale histogram of each non-dynamic area in the indoor image at each moment, and recording the reciprocal of the standard deviation as the overall contrast of each non-dynamic area in the indoor image at each moment; The grayscale values ​​of all pixels at the same position in the indoor image at all times are obtained, and the grayscale values ​​of all pixels at the same position are linearly fitted using the least squares method to obtain a grayscale fitting line for each pixel.

5. The artificial intelligence-based building lighting optimization and energy-saving control system according to claim 1 is characterized in that: Obtaining the degree of change in natural light intensity indoors in the building based on the difference in overall contrast between the same non-dynamic area in the indoor images at all times and the slope and direction of the grayscale fitting line of all pixels in the same non-dynamic area includes: The degree of illumination change in each non-dynamic area in the indoor image is obtained based on the difference in overall contrast between the same non-dynamic area in the indoor image at all times and the slope of the grayscale fitting line of all pixels in the same non-dynamic area. The degree of change in natural light intensity in the building interior is obtained based on the degree of illumination change in each non-dynamic area in the indoor image and the slope direction of the grayscale fitting line of all pixels in the non-dynamic area.

6. The artificial intelligence-based building lighting optimization and energy-saving control system according to claim 5 is characterized in that: Obtaining the degree of illumination change in each non-dynamic area in the indoor image based on the difference between the overall contrasts of the same non-dynamic area in the indoor image at all times and the slope of the grayscale fitting line of all pixels in the same non-dynamic area includes: Where, Indicates the In the indoor image at the moment The overall contrast of the non-dynamic area, Indicates the In the indoor image at the moment The overall contrast of the non-dynamic area, represents the total number of all indoor images taken in time series, Indicates the The maximum value of the slope of the grayscale fitting line of all pixels in the non-dynamic area, represents the linear normalization function, Indicates the indoor image The degree of illumination change in a non-dynamic area.

7. The artificial intelligence-based building lighting optimization and energy-saving control system according to claim 5 is characterized in that: The method of obtaining the degree of change in natural light intensity in the building interior according to the degree of change in illumination in each non-dynamic area in the indoor image and the slope direction of the grayscale fitting line of all pixels in the non-dynamic area includes: Where, Indicates the indoor image The degree of illumination change in a non-dynamic area, Indicates the The maximum value of the slope of the grayscale fitting line of all pixels in the non-dynamic area, Indicates the absolute value symbol, represents the set of all non-dynamic regions, Indicates taking the minimum value from the set of all non-dynamic areas, Indicates the degree of change in natural light intensity in a building.

8. The artificial intelligence-based building lighting optimization and energy-saving control system according to claim 1 is characterized in that: The negative mapping of the degree of change in the intensity of natural light in the building to obtain the light intensity adjustment factor includes: Where, Indicates the degree of change in the intensity of natural light in a building. Indicates the light intensity adjustment factor.

9. The artificial intelligence-based building lighting optimization and energy-saving control system according to claim 1 is characterized in that: The adjusting the original light intensity by the light intensity adjustment factor to obtain the adjusted light intensity includes: Where, Indicates the original light intensity. Indicates adjusting the light intensity. Indicates the light intensity adjustment factor.

Images