Garbage overflow monitoring method based on image processing
Through the garbage overflow monitoring method based on image processing, the camera collects image information and image binarization technology to identify garbage overflow, solving the problems of high cost and low efficiency of traditional smart garbage cans, real-time monitoring and reminding of garbage overflow is achieved. Improved sanitation work efficiency and residents' life experience.
Patent Information
- Application Number
- CN202411429248.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-14
- Publication Date
- 2025-02-11
AI Technical Summary
The existing technology has high cost and low efficiency in monitoring and handling garbage overflows, especially in the community. Traditional smart garbage cans require the installation of sensor equipment, which is relatively high overall cost and is prone to damage.
The garbage overflow monitoring method based on image processing is adopted, image information is collected through the camera, the relationship between the garbage can and the staff is established, image binarization and pixel changes are used to identify the garbage overflow situation, and staff are reminded to clean it through text messages or WeChat.
Real-time monitoring and reminder of garbage overflow situations has been achieved, which has improved sanitation work efficiency, reduced dependence on sensor equipment, reduced overall costs, and improved residents' living experience.
Smart Images
Figure CN119418264A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of image recognition, in particular to a garbage overflow monitoring method based on image processing. Background Art
[0002] With the continuous progress of urbanization, more and more waste is generated in cities. Garbage disposal has increasingly affected people's daily lives and has become an urgent issue in the comprehensive improvement of urban environment. Generally, there are multiple large garbage bins in residential areas. Garbage collection and transportation mostly adopts fixed-time fixed-route collection and transportation methods, and garbage bins often overflow. Once the garbage overflows, it will affect the community environment and increase the workload of sanitation workers. For this reason, it is urgent to use information technology to realize the status monitoring of garbage bins in the community and improve the efficiency of sanitation work. At present, the monitoring of garbage overflow in the community is mainly through the research of smart garbage bins, and the reminder of garbage overflow is fed back through the smart devices on the garbage bins. However, smart garbage bins need to install some sensor equipment, the overall cost is relatively high, and they are prone to damage, resulting in low application efficiency and limited overall impact on garbage overflow treatment. Summary of the invention
[0003] This paper collects image information based on cameras and uses image processing to achieve real-time monitoring of garbage overflow, so that the garbage can be processed as soon as possible and the happiness of residents can be improved.
[0004] A garbage overflow monitoring method based on image processing comprises the following steps:
[0005] S1. Establish the association between staff and trash can
[0006] This article first numbers each trash can, clarifies the location of each trash can, and the staff member who cleans each trash can. Each trash can is associated with a staff member. When a message of garbage overflow is identified, the staff member is reminded to deal with the garbage in time.
[0007] S2. Judgment of working hours
[0008] Set working hours to remind staff to dispose of garbage only during working hours;
[0009] S3. Image acquisition
[0010] First, determine whether each trash can is captured by a camera. If some trash cans cannot be captured directly, adjust the position of the trash can or install a camera device. The image acquisition process is as follows:
[0011] (1) Trash bin positioning
[0012] By rotating and zooming in the camera, the trash can is positioned in the middle of the camera, and the trash can is taken as the center and the height H of the trash can is taken as the reference, and the 2H range around the trash can is selected as the trash can image collection area. If there are multiple trash cans, the trash can's horizontal position is taken as the center and the height H of the trash can is taken as the reference, and the range around the trash can is selected as the trash can collection area;
[0013] 1) Determine the trash can as the center of the image
[0014] Manually adjust the camera and set the trash can to the center of the camera. In order to cooperate with the common monitoring functions of the community camera, the screen display is based on the usual 16:9 ratio of the camera, and the trash can height H is used as a reference. Select the 16H*9H range as the camera adjustment position;
[0015] 2) Determine the image acquisition area
[0016] Only select images within the range of 2H*2H with the trash can as the center and the trash can height as the reference;
[0017] (2) Image collection around the trash can
[0018] The image acquisition method is to collect images once every 5 minutes, and 3 cycles constitute a cycle. In a cycle, if there is no garbage overflow in one photo, a new cycle will start. If there is garbage overflow in three consecutive image acquisitions, it will be considered that there is garbage overflow and a background reminder will be issued. If there is no garbage overflow in any of the three consecutive image acquisitions, a new round of image acquisition will start. First, the image without garbage overflow on that day will be collected as the basic image, which will be used as the comparison image of garbage overflow later.
[0019] S4. Image Binarization
[0020] Select red, blue and green colored trash cans, collect images of the trash cans, grayscale the collected images, and binarize the images;
[0021] S5. Garbage overflow identification
[0022] If the change in the image pixel value within the range of 2H*2H between when the trash bin is not full and when the trash bin is full increases, it is considered that the trash overflow has occurred;
[0023] S6. Spam reminder
[0024] If garbage overflow is found, information will be sent to the corresponding users via SMS or WeChat to remind the staff to clean up.
[0025] It can be applied in the identification of garbage overflow in the community, allowing staff to accurately detect the overflow of garbage bins during the working period, so that staff can clean up in time and improve the living experience of residents. Compared with the traditional garbage overflow warning, there is no need to add any hardware facilities to the garbage bin, only the image recognition algorithm is needed to remind the garbage overflow. Moreover, this paper first manually locates the garbage bin, and then uses the image recognition algorithm to detect garbage overflow based on the change of pixels. First of all, for the garbage overflow itself, it saves the algorithm research on the positioning of the garbage bin and saves the algorithm pointing out; moreover, the staff is more aware of the location of the garbage bin, which improves the efficiency of the initial positioning of the garbage bin; in addition, this paper compares the pixel changes after the image is binarized, which is different from the current garbage recognition based on artificial neural networks. The overall calculation amount is much less, which is more beneficial to the overall investment and market promotion of garbage overflow recognition. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] Figure 1 The garbage overflow monitoring process for basic image processing;
[0027] Figure 2 To determine the trash can as the center of the image;
[0028] Figure 3 Schematic diagram for determining the image acquisition area;
[0029] Figure 4 Flowchart for collecting images around the trash can;
[0030] Figure 5 Schematic diagram of image binarization;
[0031] Figure 6 Schematic diagram of binarization for garbage overflow identification. DETAILED DESCRIPTION
[0032] like Figure 1 As shown in the figure, first, a corresponding relationship between trash cans and staff is established to facilitate reminding relevant personnel to handle garbage. Secondly, in order to remind staff to handle garbage in time during work, a system work monitoring is established. This article mainly uses the local sun rising and setting time as the system working time. During the system operation, the camera obtains the image around the trash can, and after image binarization, it is determined whether there is garbage overflow. If there is garbage overflow, the staff responsible for the trash can at that location will be reminded to clean up the garbage in time in the form of text messages.
[0033] 1. Establish the relationship between staff and trash cans
[0034] In order to facilitate the staff to deal with the overflowing garbage bins in time, this article first numbers each garbage bin, clarifies the location of each garbage bin, and the staff corresponding to each garbage bin cleaning. The corresponding relationship is as follows:
[0035] Table of the relationship between staff and trash can management
[0036] Num Position Name Phone 1 Trash can in front of building X in community X Li San 1XXXXXXXXX
[0037] Num is the trash bin number, Position is the location of the trash bin, Name is the name of the staff member corresponding to the trash bin, and Phone is the telephone number of the staff member. The association relationship between each trash bin and the staff member is implemented in turn. When the garbage overflow message is identified, the staff member is reminded to deal with the garbage in time.
[0038] 2. Determination of working hours
[0039] In order to remind the staff to clean up the garbage in time and provide residents with a clean and tidy living environment, and also to avoid disturbing the staff's rest time. In addition, the garbage generated in the general community is mainly the garbage brought down by residents when they get up early to go to work. Based on this, the system startup time is set around the working hours of the community staff. Outside the working hours of the community staff, the system is in a shutdown and energy-saving state.
[0040] 3. Image Acquisition
[0041] Figure 3 As shown in the figure, first determine whether each trash can is captured by a camera. If some trash cans cannot be captured directly, adjust the position of the trash can or install camera equipment. Since this article mainly solves the problem of garbage overflow reminder, this article assumes that all trash cans can be captured by cameras. After that, locate and collect images of the area near the trash can to accurately verify whether there is garbage overflow.
[0042] (1) Trash bin positioning
[0043] Since the cameras in community applications now have a rotation function, this paper positions the trash can in the middle of the camera by rotating and zooming in the camera, and takes the trash can as the center and the height H of the trash can as the reference, and selects the 2H range around the trash can as the trash can image collection area. If there are multiple trash cans, the horizontal position of the trash can is taken as the center, and the height H of the trash can is used as the reference, and the range around the trash can is selected as the trash can collection area.
[0044] 1) Determine the trash can as the center of the image
[0045] Manually adjust the camera and set the trash can as the center of the camera. In order to cooperate with the common monitoring functions of the community camera, this article will use the center of the trash can as the screen display according to the usual 16:9 ratio of the camera. In order to better play the community monitoring function of the camera when monitoring the overflow of the trash can, the trash can height H is used as a reference and the 16H*9H range is selected as the camera adjustment position.
[0046] 2) Determine the image acquisition area
[0047] In order to find the garbage overflow more clearly and accurately, during image processing, only images within the range of 2H*2H with the garbage bin as the center and the height of the garbage bin as the reference are selected.
[0048] (2) Image collection around the trash can
[0049] Figure 4 The figure shows the surrounding environment with the trash can as the center. At the same time, in order to avoid errors caused by the flow of people, the image acquisition method in this paper is to collect images every 5 minutes, and 3 cycles are one cycle; within a cycle, if there is no garbage overflow in one photo, a new cycle will start. If there is a prompt of garbage overflow in three consecutive image acquisitions, it is considered that there is garbage overflow, and a background reminder will be issued. If there is no garbage overflow in any of the three consecutive image acquisitions, a new round of image acquisition will start. First, the staff collects images without garbage overflow on that day as the basic image, which will be used as the comparison image of garbage overflow later. At the same time, in order to avoid the influence of sun shadows, the basic image is updated every hour.
[0050] 4. Image Binarization
[0051] Figure 5 As shown in the figure, since the colors of red, green, and blue trash cans are very different from the community environment, this article focuses on using the colors of objects around the trash cans to determine whether there is garbage overflow. For example, the collected image is shown in the following example:
[0052] Convert the acquired image to grayscale as shown below:
[0053] Gray(i,j)=R(i,j)*0.3+G(i,j)*0.59+B(i,j)*0.11
[0054] R(i,j), G(i,j), and B(i,j) represent the red, green, and blue primary color values of the pixel image at the i-th point on the horizontal axis and the j-th point on the vertical axis, respectively. According to the image pixel requirements, the three value ranges of R, G, and B are all [0,255], and Gray(i,j) is the grayscale value of the corresponding pixel. Since the value range of Gray(i,j) is [0,255], the image is binarized with 128 as the threshold, as shown below:
[0055]
[0056] 5. Garbage overflow identification
[0057] Figure 6 As shown, if the trash can is not full, the change in the image pixel value within the 2H*2H range is extremely small. When the trash can is full, the change in the image pixel value within the 2H*2H range will be large. We use this as a basis to determine whether the trash can is overflowing.
[0058] For the convenience of calculation, the 2H*2H area of the garbage bin cycle is uniformly divided into 40*40 areas, and the size of each area is h*h
[0059]
[0060] The pixel value of each area is calculated by comparing the number of 0 and 1 pixels in the area after binarization. If there are more 0 pixels, the area is 0; if there are more 1 pixels, the area is 1.
[0061] if(P(x,y)Num_0(0)〉P(x,y)Num_1(1)), then P(x,y)=0;
[0062] if(P(x,y)Num_0(0)<P(x,y)Num_1(1)), then P(x,y)=1;
[0063] Among them, P(x, y)Num_0(0) is the number of 0s in the coordinate range (x, y); P(x, y)Num_1(1) is the number of 1s in the coordinate range (x, y).
[0064] Figure 6 As shown below:
[0065] Num_0(0)=74
[0066] Num_1(1)=7
[0067] Since Num_0(0)〉Num_1(1), the whole image is adjusted to a blank frame. When there is garbage overflow, the whole original blank area will be filled with dark color, resulting in a large change in the image after binarization in the whole 40*40 area compared with the basic image. According to the fact that garbage overflow is mainly in the upper and bottom areas of the garbage bin, if the pixel value in each range of the 40*40 range changes, that is, from 0 to 1, Change(0->1); or from 1 to 0, Change(1->0), then the number of changed pixel values is recorded Sum=S1(Change(0->1))+S2(Change(1->0)). When the overall pixel value change range exceeds 2% of the 40*40 binary pixel value, garbage overflow is considered to occur. S1(Change(0->1)) is the number of changes from 0 to 1 in the whole area; S2(Change(1->0)) is the number of changes from 1 to 0 in the whole area. That is, if Sum>40*40*0.02, that is, Sum is greater than 32, it is considered that garbage overflow has occurred.
[0068] 6. Spam reminder
[0069] If garbage overflow is found, information will be sent to the corresponding users in the form of text messages or WeChat based on the information of each garbage bin manager, and the staff can quickly clean up the garbage.
[0070] Glossary
[0071] Image processing, also known as image processing, is a technology that uses computers to process images to achieve the desired results. It originated in the 1920s and is generally digital image processing. The main contents of image processing technology include three parts: image compression, enhancement and restoration, and matching description and recognition. Common processing includes image digitization, image coding, image enhancement, image restoration, image segmentation and image analysis. Image processing is the use of computers to process image information to meet human visual psychology or application needs. It is widely used and is mostly used in surveying and mapping, atmospheric science, astronomy, image beautification, and improving image recognition.
[0072] Garbage overflow refers to the phenomenon that the garbage in a garbage bin or garbage disposal facility exceeds the designed capacity, causing the garbage to overflow the garbage bin or facility.
Claims
1. A garbage overflow monitoring method based on image processing, characterized in that The steps include: S1. Establish the association between staff and trash can First, number each trash can, identify the location of each trash can, and the staff member who cleans each trash can. Each trash can is associated with a staff member. When a message about garbage overflow is identified, the staff member is reminded to deal with the garbage in a timely manner. S2. Judgment of working hours Set working hours to remind staff to dispose of garbage only during working hours; S3. Image acquisition First, determine whether each trash can is captured by a camera. If some trash cans cannot be captured directly, adjust the position of the trash can or install a camera device. The image acquisition process is as follows: (1) Trash bin positioning By rotating and zooming in the camera, the trash can is positioned in the middle of the camera, and the trash can is taken as the center and the height H of the trash can is taken as the reference, and the 2H range around the trash can is selected as the trash can image collection area. If there are multiple trash cans, the trash can's horizontal position is taken as the center and the height H of the trash can is taken as the reference, and the range around the trash can is selected as the trash can collection area; 1) Determine the trash can as the center of the image Manually adjust the camera and set the trash can to the center of the camera. In order to cooperate with the common monitoring functions of the community camera, the screen display is based on the usual 16:9 ratio of the camera, and the trash can height H is used as a reference. Select the 16H*9H range as the camera adjustment position; 2) Determine the image acquisition area Only select images within the range of 2H*2H with the trash can as the center and the trash can height as the reference; (2) Image collection around the trash can The image acquisition method is to collect images once every 5 minutes, and one cycle is 3 times. In one cycle, if there is no garbage overflow in one photo, a new cycle will start. If there is garbage overflow in three consecutive image acquisitions, it will be considered that there is garbage overflow and a background reminder will be issued. If no garbage overflow is found in any of the three consecutive image acquisitions, a new round of image acquisition will begin. First, the image without garbage overflow on that day will be collected as the basic image, which will be used as the comparison image for garbage overflow later. S4. Image Binarization Select red, blue and green colored trash cans, collect images of the trash cans, grayscale the collected images, and binarize the images; S5. Garbage overflow identification If the change in the image pixel value within the range of 2H*2H between when the trash bin is not full and when the trash bin is full increases, it is considered that the trash overflow has occurred; S6. Spam reminder If garbage overflow is found, information will be sent to the corresponding users via SMS or WeChat to remind the staff to clean up.