System and method for automatically adjusting water quantity of sprinkler nozzle based on machine vision

Abstract

The invention belongs to the technical field of intelligent watering cart spraying systems, and discloses a watering cart nozzle water volume automatic adjusting system and method based on machine vision, and the system comprises an image acquisition device, an image processing device and a watering control device. The sprinkler nozzle water volume automatic adjusting method comprises the steps that firstly, a road image is collected, and the road width is obtained according to the image; then putting the image into a neural network for training to obtain a target category, a speed and a distance; finally, the water volume of the sprinkler nozzle is automatically adjusted by controlling the water pump pressure of the sprinkler and the start-stop time of the water pump. The road distance is judged by detecting the center point of the edge of the road, so that the water spraying amount of the three-bend front flushing spray head is controlled, and water can be effectively saved; the neural network is combined with the binocular camera distance measurement module to realize speed and distance detection of pedestrians and non-motor vehicles, so that the water spraying amount of the three-bend forward-flushing spray head is regulated and controlled, and real-time avoidance of the pedestrians and the non-motor vehicles is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent sprinkler truck spraying systems, and in particular to a system and method for automatically adjusting the water volume of a sprinkler truck nozzle based on machine vision. Background Art

[0002] As a type of sanitation vehicle, sprinkler trucks are used for cleaning urban roads, parks, construction sites and other places. Sprinkler trucks have the functions of sprinkling water to reduce dust, watering greenery, and washing the ground. They can effectively improve environmental quality and maintain environmental hygiene. They are an important tool for urban management and environmental protection work.

[0003] At present, the high-pressure water guns on sprinkler trucks are mostly connected to three-bend forward-spraying nozzles. Compared with traditional direct-spraying nozzles, the three-bend design of the three-bend forward-spraying nozzles makes the water flow at a softer angle when spraying, reducing the impact of the water flow on the ground, which helps to protect the ground vegetation and road surface. However, since the three-bend forward-spraying nozzles of sprinkler trucks need to spray the entire road surface, the flow rate often needs to be set to a larger fixed value and will not change, which will cause a waste of water resources; in addition, since the nozzle diameter of the three-bend forward-spraying nozzle is smaller, the pressure of the water sprayed is higher than that of the duckbill nozzle at the same flow rate, which will not only splash on pedestrians on the roadside, but also affect passing non-motorized vehicles, and cannot achieve the intelligent spraying effect. Summary of the invention

[0004] The present invention is intended to provide a system and method for automatically adjusting the water volume of a sprinkler head based on machine vision, using an image acquisition device to collect original road images, using an image processing device to calculate the center point of the road edge and the target category, speed and distance, and using a sprinkler control device to control the time interval and distance of sprinkler watering. The problem in the background technology is solved.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] A machine vision-based automatic water volume adjustment system for a sprinkler head, comprising an image acquisition device, an image processing device, and a sprinkler control device;

[0007] The image acquisition device is used to collect the original road image, transfer the original image to the image processing device, and convert the original image into a depth map;

[0008] The image processing device is used to process the received original image, calculate the center point of the road edge and the target category, speed and distance, and adjust the watering situation;

[0009] The watering control device is used to control the watering time interval and watering distance to realize intelligent watering.

[0010] Furthermore, the image acquisition device includes a binocular camera with a camera bracket, the binocular camera includes a left camera head and a right camera head, and the left camera head and the right camera head are connected at the same horizontal height on the head of the sprinkler truck.

[0011] Furthermore, the image acquisition device is a Raspberry Pi.

[0012] Furthermore, the sprinkler control device includes a water pump and a three-bend forward-flushing nozzle, the three-bend forward-flushing nozzle is connected to the water pump, the water pump is electrically connected to the Raspberry Pi, and the Raspberry Pi adjusts the sprinkler distance by controlling the pressure of the water pump, and controls the sprinkler time interval by controlling the start and stop of the water pump.

[0013] Furthermore, in the process of collecting the original road image, the calibration method of the binocular camera is as follows: the Zhang Zhengyou calibration method is used to solve the intrinsic parameter matrix A, the extrinsic parameter matrix [RT] is obtained through the intrinsic parameter matrix and the homography matrix H, and the radial distortion and tangential distortion are corrected by the distortion correction model, thereby obtaining the radial distortion parameters k1, k2, k3 and tangential distortion parameters p1, p2 of the binocular camera, and obtaining the forward projection mapping between the world coordinate system and the image coordinate system:

[0014]

[0015] Where s is the scale factor from the world coordinate system to the image coordinate system, is the image coordinate.

[0016] Furthermore, the method for converting the original image into a depth map by the binocular camera is as follows: the left camera and the right camera take the same picture of the same target, find the corresponding matching point of each pixel in the left camera image in the right camera image, and then calculate the disparity by comparing the position difference of the corresponding pixel points in the left camera and right camera images, thereby obtaining the depth information of the target.

[0017] Furthermore, the method for the image processing device to calculate the center point of the road edge is: using the obtained external parameters and distortion parameters of the binocular camera to perform distortion correction and epipolar line correction on the image, combining the binocular camera to obtain a depth map of the road image, and obtaining the coordinates of the center point of the road edge through binarization, corrosion, expansion and finding the maximum color block image operation.

[0018] Furthermore, the image processing device uses a neural network to calculate the target category and distance, wherein the neural network includes a backbone network, a fusion network and a head network; the backbone network uses convolution kernels of different sizes to extract feature layers of different scales from the input data; the fusion network performs diversified fusion of these feature layers of different scales to generate a new fusion feature map, and inputs the combined feature map into the head network of target detection. The decision unit of the head network calculates the target category and distance based on these fusion feature maps, predicts the coordinates of the target, and generates a prediction box.

[0019] Furthermore, the backbone network is a feature extraction network, which adopts a CSPNet structure, and reduces the amount of calculation by dividing the feature map into two parts and processing them in different paths;

[0020] The fusion network uses a feature pyramid structure to upsample the low-resolution feature map and fuse it with the high-resolution feature map in a bottom-up manner, enriching the semantic information and detail information of the feature to identify and locate the target;

[0021] The head network converts the feature map into the category to which each pixel belongs through several fully connected layers and generates a prediction box.

[0022] Furthermore, the position coordinates of the target predicted by the head network include coordinates of a bounding box, a category probability, and a confidence level.

[0023] Furthermore, the method for calculating the target speed by the image processing device is: first read the video image frame by frame, then obtain the coordinates of the target frame through ByteTrack tracking, calculate the height and distance of the target frame, and calculate the target speed according to the frame rate and distance.

[0024] Furthermore, the watering control device controls the watering time interval to be:

[0025]

[0026] In the formula, t is the time to control the watering, specifically, the amount of watering is reduced after t seconds, d l v is the distance between pedestrians or non-motorized vehicles and the front end of the sprinkler truck along the road direction, m is the relative speed between pedestrians or non-motorized vehicles and the sprinkler truck; t x is a correction factor, which includes visual recognition time, system judgment time and water spraying braking time.

[0027] The above-mentioned method of automatically adjusting the water volume of a sprinkler nozzle by a sprinkler nozzle water volume automatic adjustment system based on machine vision first uses the Zhang Zhengyou calibration method to calibrate the binocular camera and collect road images, and then performs road edge detection on the collected images to obtain the center point of the road edge, thereby obtaining the road width; then the collected images are put into a neural network for training to obtain the target category, speed and distance; finally, the data is processed by Raspberry Pi, and the automatic adjustment of the water volume of the sprinkler nozzle is achieved by controlling the water pump pressure and the start and stop time of the water pump.

[0028] The beneficial effects of the technical solution are:

[0029] 1. The present invention provides a system and method for automatic water volume adjustment of a sprinkler nozzle based on machine vision. The image acquisition device adopts a binocular camera, and the left and right cameras are located at the same horizontal height, which ensures the accuracy and consistency of image acquisition and provides a high-quality data basis for subsequent image processing; the image acquisition device adopts Raspberry Pi, which not only has powerful computing power, but also has good scalability and programmability, making the system more flexible and easy to maintain; by controlling the pressure and start and stop time of the water pump, the watering distance and watering time interval can be accurately adjusted to meet the watering needs under different road and weather conditions.

[0030] 2. The present invention provides a system and method for automatic water volume adjustment of sprinkler nozzles based on machine vision. The Zhang Zhengyou calibration method is used to solve the internal parameter matrix, and the radial and tangential distortions are corrected by the distortion correction model. The images collected by the binocular camera have good authenticity and accuracy. The neural network is used to calculate the target category and distance, which improves the accuracy and robustness of target recognition and provides rich decision-making basis for sprinkler control. The backbone network adopts the CSPNet structure, which effectively reduces the amount of calculation and improves the real-time and energy efficiency of the system. The fusion network enriches the semantic information and detail information of the features through the feature pyramid structure, and improves the accuracy of target recognition and positioning. The head network converts the feature map into the category and prediction box of each pixel through the fully connected layer, which provides accurate target position information for sprinkler control. The sprinkler control device flexibly adjusts the sprinkler time interval according to the relative position and speed of pedestrians or non-motor vehicles and the sprinkler truck, as well as the correction coefficient, to ensure the safety and effectiveness of the sprinkler operation.

[0031] In summary, the present invention can effectively save water by detecting the center point of the road edge to determine the road distance, thereby controlling the watering amount of the three-bend forward-rushing sprinkler. The neural network is combined with the binocular camera ranging module to realize the speed and distance detection of the target (pedestrians, non-motorized vehicles), thereby adjusting the watering amount of the three-bend forward-rushing sprinkler to achieve real-time pedestrian and non-motorized vehicle avoidance. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a flow chart of a system for automatically adjusting the water volume of a sprinkler head based on machine vision according to the present invention;

[0033] Figure 2 This is a flow chart of binocular camera ranging in the present invention;

[0034] Figure 3 A flow chart of road edge distance measurement of the present invention;

[0035] Figure 4 This is a schematic diagram of the network structure of the present invention;

[0036] Figure 5 It is a diagram of the speed and distance detection results of pedestrians and non-motor vehicles of the present invention;

[0037] Figure 6 This is a schematic diagram of distance judgment according to the present invention. DETAILED DESCRIPTION

[0038] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments:

[0039] A machine vision-based automatic water volume adjustment system for a sprinkler head, comprising an image acquisition device, an image processing device, and a sprinkler control device;

[0040] The image acquisition device includes a binocular camera with a camera bracket, the binocular camera includes a left camera and a right camera, and the left camera and the right camera are connected to the same horizontal height on the head of the sprinkler truck, the binocular camera is used to collect the original road image, and transmit the original image to the image processing device, and convert the original image into a depth map;

[0041] The image acquisition device is a Raspberry Pi, which is used to process the received raw images, calculate the center point of the road edge and the target category, speed and distance, and adjust the watering situation;

[0042] The sprinkler control device includes a water pump and a three-bend forward-flushing nozzle. The three-bend forward-flushing nozzle is connected to the water pump. The water pump is electrically connected to the Raspberry Pi. The Raspberry Pi adjusts the sprinkler distance by controlling the pressure of the water pump to achieve intelligent sprinkler.

[0043] like Figure 1As shown, a method for automatically adjusting the water volume of a sprinkler nozzle by an automatic water volume adjustment system of a sprinkler nozzle based on machine vision, firstly, a binocular camera is used to capture a road image and convert it into a depth map, and the center point of the road edge is extracted from the depth map to obtain the distance from the road edge; then a neural network is used to obtain pedestrian and non-motor vehicle information, and obtain their distance from the front end of the vehicle, and then combined with the distance from the vehicle to the road edge to obtain the distance along the road direction, the speed is estimated by the distance on the two frames of images, so as to obtain the meeting time, and the water spraying volume is controlled by controlling the water pump pressure of the sprinkler, so as to achieve the purpose of real-time detection and control.

[0044] like Figure 2 As shown in the figure, target detection and binocular ranging based on neural network are as follows:

[0045] Calibrate the stereo camera on the road image: Use Zhang Zhengyou's calibration method to solve the intrinsic parameter matrix A, and obtain the extrinsic parameter matrix [RT] through the intrinsic parameter matrix and the homography matrix H. Then use the distortion correction model to correct the radial distortion and tangential distortion, so as to obtain the radial distortion parameters k1, k2, k3 and tangential distortion parameters p1, p 2， Then complete the forward projection mapping between the world coordinate system and the image coordinate system:

[0046]

[0047] Where s is the scale factor from the world coordinate system to the image coordinate system, is the image coordinate.

[0048] The principle of binocular camera distance measurement is as follows: the binocular camera consists of two cameras, the left camera and the right camera, which are placed within a certain distance, which is called the baseline. The two cameras shoot the same scene to generate different images, the left camera image is L, and the right camera image is R.

[0049] Due to the different placement positions, there is a difference in the viewing angles of the two cameras, and the corresponding pixel positions of the same object on L and R are also different. This difference is called parallax, so the object distance z is obtained as:

[0050]

[0051] Where b is the baseline distance between the two cameras, f is the focal length of the camera, and x r and x l is the horizontal coordinate of the feature point on L and R. The target detection steps are as follows:

[0052] Collect road pictures, create training set labels and conduct training, obtain the trained weight file, evaluate the weight file parameters to obtain the best weight file, and import the trained weight file into the neural network for prediction.

[0053] like Figure 3 As shown in the figure, the road edge distance measurement based on machine vision is as follows:

[0054] First, the captured road image is binarized, and the global threshold method is used to set pixels greater than the threshold to white, and pixels less than the threshold to black. The formula is as follows:

[0055]

[0056] Where f(x, y) is the pixel value after operation, g(x, y) is the original pixel value, and T is the threshold.

[0057] The binarized image is preprocessed and opened; the convolution kernel slides along the image. If all pixel values ​​of the original image corresponding to the convolution kernel are 1, the central element retains the original pixel value, otherwise it is 0, that is, the image is corroded.

[0058] Then, the expansion operation is performed, that is, as long as one of the pixel values ​​of the original image corresponding to the convolution kernel is 1, the pixel value of the central element is 1; the opening operation can eliminate noise, fill small holes in the image, and make the depth map smoother. After preprocessing, the largest color block in the image is found to obtain the road edge information, and all pixel values ​​except the road edge are set to 0 to calculate the coordinates of the road center point.

[0059] Finally, calculate the distance between the sprinkler truck and the center point of the road edge:

[0060] The conversion formula from pixel to image is:

[0061]

[0062] Where (u, v) is the pixel coordinate, d x , d y Represents the length of a single pixel in the x and y directions respectively.

[0063] So the coordinates of the pixel point in the camera coordinate system are:

[0064]

[0065] Where D(u, v) is the depth information of the pixel point (u, v), (f x , f y , u0, v0) are the camera intrinsic parameters obtained by camera calibration.

[0066] Then, transform the coordinates in the camera coordinate system to the world coordinate system:

[0067] PW =[RT]*P C

[0068] Where [RT] is the camera extrinsic parameter obtained by camera calibration, P W (X W , Y W , Z W ) is its coordinate in the world coordinate system, P C are the coordinates in the camera coordinate system.

[0069] Finally, the actual distance between the camera and the center point of the road edge is:

[0070]

[0071] If the camera height is h, the horizontal distance is:

[0072]

[0073] like Figure 4 As shown, the neural network includes a backbone network, a fusion network and a head network; its specific structure is:

[0074] The backbone network is a feature extraction network that uses a CSPNet structure. It can effectively reduce the amount of computation by dividing the feature map into two parts and processing them in different paths.

[0075] The fusion network uses a feature pyramid structure to upsample low-resolution feature maps and fuse them with high-resolution feature maps in a bottom-up manner, enriching the semantic information and detail information of the features, so that the subsequent detection head can more accurately identify and locate the target;

[0076] The head network converts the feature map into the category to which each pixel belongs through several fully connected layers and generates a prediction box.

[0077] like Figure 5 The figure shows the speed and distance detection results of pedestrians and non-motorized vehicles. The principle of the speed detection module is as follows:

[0078] First, read the frame rate of the video and read the video image frame by frame; obtain the coordinates (X1, y1), (x2, y2) of the target frame through ByteTrack tracking, and calculate the height of the target frame:

[0079] H=y2-y1

[0080] Estimate the distance of the target box by its height:

[0081]

[0082] In the formula, H ris the actual height of the object. Since pedestrians and non-motorized vehicles are detected, H r Set to a certain value, f is the focal length of the camera. Since the coordinate frame should completely surround the target object, the coordinate frame height should be slightly higher than the object height, so the corrected target frame height H* is used.

[0083] The speed measurement is achieved by comparing the distance change of the object between different frames. First, the distance of the object in each frame is recorded, and then the moving speed of the object is calculated by the frame interval:

[0084]

[0085] Where D2 is the target frame distance of the current frame, D1 is the target frame distance of the previous frame, and t is the time interval, which is calculated by the frame rate.

[0086] like Figure 6 As shown, the distance judgment principle is as follows:

[0087]

[0088] Where, d l is the distance between the pedestrian or non-motorized vehicle and the front end of the sprinkler truck along the road direction, D is the distance between the camera and the target, h is the camera height, and L is the lateral distance between the camera and the road edge.

[0089] According to the target distance and speed, the controlled watering time can be obtained as:

[0090]

[0091] In the formula, t is the time to control the watering, that is, the amount of watering is reduced after t seconds, d l v is the distance between pedestrians or non-motorized vehicles and the front end of the sprinkler truck along the road direction, m is the relative speed between pedestrians or non-motorized vehicles and the sprinkler truck. x is a correction factor, which includes visual recognition time, system judgment time and water spraying braking time.

[0092] The amount of water sprayed can be controlled by adjusting the pressure and speed of the water pump. The water pump uses mechanical force to pump water from a low place to a high place or transport it over a long distance. Its core components are the pump body and the impeller. When the water pump is working, the impeller rotates to generate centrifugal force, sucking water from the pump inlet and accelerating it, and then discharging it through the outlet.

[0093] Among them, the principle of controlling the water spraying amount by controlling the water pump pressure is:

[0094] The pressure of the water pump can be controlled by adjusting the outlet valve of the pump. When the outlet valve is opened to a small degree, the water flow rate increases and the pressure increases, which will cause the water column to be sprayed farther and more concentrated, thereby reducing the amount of water sprayed per unit area. On the contrary, if the valve opening is increased, the pressure decreases, the water flow rate slows down, and the amount of water sprayed will be greater, covering a wider area.

[0095] The nozzle flow rate is controlled by controlling the water pump pressure. The specific formula is as follows:

[0096]

[0097] Where q is the nozzle flow rate, P is the nozzle working pressure, K is the nozzle flow coefficient, g is the gravitational acceleration, and h is the height between the nozzle and the ground.

[0098] In summary, the present invention provides a system and method for automatically adjusting the water volume of a sprinkler head based on machine vision. By detecting the center point of the road edge to determine the road distance, the water volume of the three-bend forward-rushing sprinkler head can be controlled, which can effectively save water. The neural network is combined with the binocular camera ranging module to realize the speed and distance detection of the target (pedestrians, non-motorized vehicles), thereby adjusting the water volume of the three-bend forward-rushing sprinkler head and realizing real-time pedestrian and non-motorized vehicle avoidance.

[0099] The above is only an embodiment of the present invention, and the common knowledge such as the known specific technical solutions or characteristics in the solution is not described in detail here. It should be pointed out that for those skilled in the art, several modifications and improvements can be made without departing from the technical solution of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A sprinkler head water volume automatic adjustment system based on machine vision, characterized in that: It includes an image acquisition device, an image processing device, and a watering control device; The image acquisition device is used to collect the original road image, transfer the original image to the image processing device, and convert the original image into a depth map; The image processing device is used to process the received original image, calculate the center point of the road edge and the target category, speed and distance, and adjust the watering situation; The watering control device is used to control the watering time interval and watering distance to realize intelligent watering.

2. According to the machine vision-based automatic water volume adjustment system for sprinkler heads of a water truck according to claim 1, it is characterized in that: The image acquisition device comprises a binocular camera with a camera bracket, the binocular camera comprises a left camera head and a right camera head, and the left camera head and the right camera head are connected at the same horizontal height on the head of the sprinkler truck.

3. According to claim 2, a sprinkler head water volume automatic adjustment system based on machine vision is characterized in that: The image acquisition device is a Raspberry Pi.

4. According to the machine vision-based automatic water volume adjustment system for sprinkler heads of water trucks as described in claim 3, it is characterized in that: The sprinkler control device includes a water pump and a three-bend forward-flushing nozzle, the three-bend forward-flushing nozzle is connected to the water pump, the water pump is electrically connected to the Raspberry Pi, and the Raspberry Pi adjusts the sprinkler distance by controlling the pressure of the water pump, and controls the sprinkler time interval by controlling the start and stop of the water pump.

5. The automatic water volume adjustment system for sprinkler heads based on machine vision according to claim 4 is characterized in that: In the process of collecting original road images, the calibration method of the binocular camera is as follows: the Zhang Zhengyou calibration method is used to solve the intrinsic parameter matrix A, the extrinsic parameter matrix [RT] is obtained through the intrinsic parameter matrix and the homography matrix H, and the radial distortion and tangential distortion are corrected using the distortion correction model, thereby obtaining the radial distortion parameters k1, k2, k3 and tangential distortion parameters p1, p2 of the binocular camera, and obtaining the forward projection mapping between the world coordinate system and the image coordinate system: Where s is the scale factor from the world coordinate system to the image coordinate system, is the image coordinate.

6. The automatic water volume adjustment system for sprinkler heads based on machine vision according to claim 5 is characterized in that: The method of converting the original image into a depth map by a binocular camera is as follows: the left camera and the right camera take the same picture of the same target, find the corresponding matching point of each pixel in the left camera image in the right camera image, and then calculate the disparity by comparing the position difference of the corresponding pixel points in the left camera and right camera images, so as to obtain the depth information of the target.

7. The automatic water volume adjustment system for sprinkler heads based on machine vision according to claim 6 is characterized in that: The method for calculating the center point of the road edge by the image processing device is as follows: using the obtained extrinsic parameters and distortion parameters of the binocular camera to perform distortion correction and epipolar line correction on the image, combining the binocular camera to obtain a depth map of the road image, and obtaining the coordinates of the center point of the road edge through binarization, corrosion, expansion and searching for the largest color block image operation.

8. The automatic water volume adjustment system for sprinkler heads based on machine vision according to claim 7 is characterized in that: The image processing device uses a neural network to calculate the target category and distance, wherein the neural network includes a backbone network, a fusion network and a head network; the backbone network uses convolution kernels of different sizes to extract feature layers of different scales from input data; the fusion network performs diversified fusion of these feature layers of different scales to generate a new fusion feature map, and inputs the combined feature map into the head network of target detection. The decision unit of the head network calculates the target category and distance based on these fusion feature maps, predicts the coordinates of the target, and generates a prediction box.

9. The automatic water volume adjustment system for sprinkler heads based on machine vision according to claim 8 is characterized in that: The backbone network is a feature extraction network, which adopts the CSPNet structure and reduces the amount of calculation by dividing the feature map into two parts and processing them in different paths; The fusion network uses a feature pyramid structure to upsample the low-resolution feature map and fuse it with the high-resolution feature map in a bottom-up manner, enriching the semantic information and detail information of the feature to identify and locate the target; The head network converts the feature map into the category to which each pixel belongs through several fully connected layers and generates a prediction box.

10. The automatic water volume adjustment system for sprinkler heads based on machine vision according to claim 8 is characterized in that: The position coordinates of the target predicted by the head network include the coordinates of the bounding box, the category probability and the confidence.

11. The automatic water volume adjustment system for sprinkler heads based on machine vision according to claim 8 is characterized in that: The method for calculating the target speed by the image processing device is: first read the video image frame by frame, then obtain the coordinates of the target frame through ByteTrack tracking, calculate the height and distance of the target frame, and calculate the target speed according to the frame rate and distance.

12. The automatic water volume adjustment system for sprinkler heads based on machine vision according to claim 11 is characterized in that: The watering control device controls the watering time interval to be: In the formula, t is the time to control the watering, specifically, the amount of watering is reduced after t seconds, d l v is the distance between pedestrians or non-motorized vehicles and the front end of the sprinkler truck along the road direction, m is the relative speed between pedestrians or non-motorized vehicles and the sprinkler truck; t x is a correction factor, which includes visual recognition time, system judgment time and water spraying braking time.

13. The method for automatically adjusting the water volume of a sprinkler head by using a sprinkler head water volume automatic adjustment system based on machine vision according to claim 12, characterized in that: First, the binocular camera is calibrated using Zhang Zhengyou's calibration method and road images are collected. The road edge is detected on the collected images to obtain the center point of the road edge and thus the road width. The collected images are then put into a neural network for training to obtain the target category, speed and distance. Finally, the data is processed by Raspberry Pi, and the water volume of the sprinkler nozzle is automatically adjusted by controlling the water pump pressure and the start and stop time of the sprinkler pump.

Images