Smart inspection robot for cattle farms
By combining cattle distribution and environmental detection in the cattle farm intelligent inspection robot, the precise positioning and targeted spraying of ammonia concentration in the cattle house is achieved, which solves the problems of inaccurate detection and large spraying range, and improves the cattle house Environmental health and growth conditions.
Patent Information
- Application Number
- CN202411399866.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-09
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-10-09
AI Technical Summary
The existing intelligent inspection robot in cattle farms has inaccurate detection of ammonia concentration in closed cattle houses, resulting in a large and wasteful spraying range of ammonia remover, affecting the growth and development of cattle.
Using a smart patrol robot including a collection unit, an identification unit, an ammonia detection unit and an eye detection unit, the number and distribution of cattle can be identified accurately and the area with high ammonia concentration is sprayed for detection and ammonia removal agent, combined with feces and pads Grass area inspection to achieve targetedness and accuracy.
It improves the accuracy of ammonia concentration detection and the targeted spraying, reduces the waste of ammonia eliminators, protects the growth and development of cattle, and promptly detects and tracks cattle with red eye disease, reducing the spread of the disease.
Smart Images

Figure CN118893642B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of cruise robots, and in particular to a smart inspection robot for cattle farms. Background Art
[0002] The smart inspection robot for cattle farms usually patrols and monitors cattle farms, especially monitors environmental factors in cattle farms, and issues alarms when there are abnormal conditions. Ammonia is produced in cattle farms. It is a colorless gas with a strong irritating odor. It is one of the harmful gases in cattle sheds. Its main sources are the feces and urine of cattle, gastrointestinal digestion, and the accumulation of feces, urine, feed residues, bedding and other organic matter. Ammonia seriously affects the growth and development of cattle and harms the health of their eyes and respiratory tract. Under the current circumstances, most cattle sheds need to be sealed in winter to ensure the temperature in the shed. When the cattle shed is closed, the ammonia content increases, which is not conducive to the growth of cattle.
[0003] In the case of non-isolated breeding, the uneven distribution of cattle leads to different ammonia content in different parts. When spraying ammonia removers in the prior art, the concentration of ammonia is usually detected at fixed points, and the concentration detection is inaccurate. After the detection, large-scale spraying is carried out, which not only wastes the agent, but also is not conducive to the growth of cattle. Summary of the invention
[0004] In order to solve the problem of inaccurate detection of ammonia concentration at fixed points in closed cattle sheds and large-scale spraying of ammonia removers after detection, resulting in waste of agents, the present invention provides a cattle farm intelligent inspection robot, the robot comprising:
[0005] Collection unit: used to collect cattle farm data based on collection equipment;
[0006] Recognition unit: used for obtaining an environment image of a cattle farm and an eye image of a cattle based on the cattle farm data;
[0007] Ammonia detection unit: used to obtain a detection area based on the environmental image, control the ammonia detection device to detect the detection area, and obtain an ammonia detection result; based on the ammonia detection result, control the ammonia spraying device to spray an ammonia remover on the cattle farm;
[0008] Eye detection unit: used to detect the cattle based on the eye image, obtain eye detection results, and transmit the eye detection results to the user end.
[0009] The principle of the present invention is as follows: by identifying the number of cattle to obtain the distribution of cattle, obtain areas with many cattle, detect the ammonia concentration in the places where the cattle gather, detect whether it affects the growth and development of cattle, detect areas without cattle but with much feces and straw, detect whether their living environment produces a large amount of ammonia, and combine the two to detect, the detection point is more flexible, more targeted, and the detection concentration is more accurate, and then the concentration of the ammonia remover is configured according to the ammonia concentration, and the detection area is sprayed, and accurate spraying is performed on a small range to reduce the waste of agents.
[0010] Furthermore, the identification unit specifically includes:
[0011] A first recognition unit is used to obtain a first image based on the cattle farm data, recognize the first image based on a recognition model to obtain a first recognition result, divide the cattle farm based on the first recognition result to obtain a plurality of sub-areas, and obtain the environment image based on all the sub-areas;
[0012] The second recognition unit is used to recognize the environmental image based on the recognition model to obtain a cow image, and recognize the cow image based on the recognition model to obtain a second recognition result; if the second recognition result is successful, then the eye image is obtained based on the second recognition result; if the second recognition result is a failure, then a number of cows to be tested are obtained based on the second recognition result, and the images to be tested of the cows to be tested are obtained, and the images to be tested are recognized based on the recognition model to obtain facial features; the number of features of the facial features is obtained, and a first coordinate is obtained based on the number of features and the facial features, the robot moves to the first coordinate, and performs image acquisition based on the acquisition device to obtain a second image, and the cow image is updated to the second image.
[0013] Cows are identified according to the recognition model, and then the cattle farm is divided into different areas according to the distribution of cattle, so as to facilitate subsequent targeted detection; considering that the cattle shed is closed in winter, the environment is humid and poorly ventilated, toxic and harmful gases such as ammonia, carbon dioxide and hydrogen sulfide will be produced to irritate the eyes of cattle, causing cattle to contract infectious keratoconjunctivitis, commonly known as pink eye, with symptoms including conjunctival congestion and corneal opacity, etc. This disease is an acute infectious disease and needs to be treated and isolated as soon as possible. The present invention obtains the eye area of the identified cattle, adjusts the position of the robot in real time according to its facial features and its number, and obtains the eye area more accurately and clearly.
[0014] Furthermore, the ammonia detection unit specifically includes:
[0015] The first ammonia unit is used to obtain the number of cattle in each of the sub-areas based on the environmental image, and to sort all the sub-areas based on the number of cattle to obtain an area list; obtain the sub-areas within a preset ranking range in the area list to obtain a first detection area; obtain the sub-area where the number of cattle is equal to 0, obtain the image to be tested, obtain the feces area and the bedding area based on the image to be tested, obtain the first area and the second area of the feces area and the bedding area respectively, and obtain the second detection area based on the first area and the second area; obtain the detection area based on the first detection area and the second detection area;
[0016] A second ammonia unit: used to obtain a second coordinate based on the detection area, the robot moves to the second coordinate, controls the ammonia detection device to detect the detection area, and obtains an ammonia detection result;
[0017] The third ammonia unit is used to obtain a first concentration of ammonia based on the ammonia detection result, and determine whether the first concentration is greater than a preset concentration. If so, a second concentration of ammonia removal agent is obtained based on the first concentration. Based on the second concentration, the robot controls the ammonia spraying equipment to spray the ammonia removal agent on the detection area.
[0018] Furthermore, the eye detection unit is specifically used to: divide the eye image into several sub-images, obtain the pixel value of each sub-image, determine whether the pixel value is within a preset pixel range, if so, obtain the sub-image to obtain a dangerous area, obtain the number of dangers in the dangerous area, determine whether the number of dangers is greater than a preset number, if greater than the preset number, obtain sick cows based on the eye image, track the sick cows, obtain the eye detection result, and transmit the eye detection result to the user end.
[0019] According to the pixel values, cattle with red eye disease are obtained and tracked to reduce the spread of red eye disease and ensure the normal growth of cattle. Environmental problems in the cowshed are also reflected and returned to the user end for further processing.
[0020] Furthermore, the robot also includes a data unit, and the data unit specifically includes:
[0021] The first data unit is used to store the cattle number in the database;
[0022] A second data unit is used for collecting the first audio data of the cattle by the audio device based on the number, and storing the first audio data and the corresponding relationship between the number and the first audio data in the database;
[0023] The third data unit is used to divide the first audio data into several classified audios based on preset classifications, and store the corresponding relationship between the first audio data and all the classified audios, all the classified audios and the category of each classified audio in the database;
[0024] The fourth data unit is used to obtain the genealogical relationship between the cattle, and store the genealogical relationship and the number corresponding to the genealogical relationship in the database.
[0025] The cattle numbers, audio data, and genealogical relationships are stored to facilitate use in subsequent steps.
[0026] Genealogical relationships refer to the kinship between cattle, including close and distant relatives.
[0027] Parent-offspring relationship refers to the blood relationship between a cow and her calf.
[0028] Furthermore, the ammonia detection unit also includes:
[0029] Driving unit: used for obtaining the first number of cattle in the detection area, judging whether there is a second number in the database that has a parent-child relationship with the first number, and if so, obtaining the second audio data of the second number; based on the first preset category, obtaining the first classification audio corresponding to the second audio data, and the playback device plays the first classification audio, and after there are no cattle in the detection area, the ammonia spraying device sprays the ammonia remover on the detection area; if not, obtaining several second classification audios corresponding to the second preset category, and the playback device plays the second classification audios in sequence, and after there are no cattle in the detection area, the ammonia spraying device sprays the ammonia remover on the detection area.
[0030] Considering that the calls of cows can convey different information through changes in frequency and duration, cows can call and locate each other through calls, and the calls of each pair of cows and calves are unique, and the sounds they make are only for each other, so as to identify each other. The present invention plays the audio to move the cows away from the detection area and then sprays the ammonia remover to avoid the position of the cows, thereby reducing the situation where the agent is sprayed on the cows, which is not conducive to the growth of the cows.
[0031] Furthermore, the robot also includes:
[0032] Anomaly detection unit: used to obtain the feeding time of the cattle based on the number and the preset time range; identify the cattle image based on the recognition model, obtain the cattle head area and the cattle area, obtain the third coordinates and the fourth coordinates of the cattle head area and the cattle area respectively, judge whether the cattle has fence climbing behavior based on the third coordinates and the fourth coordinates, if so, obtain the fence climbing cattle based on the third coordinates, obtain the third number and the third audio data of the fence climbing cattle, obtain the fourth audio data corresponding to the third number in the database, judge whether the fourth audio data has a third preset category, if the third preset category exists, obtain the third classified audio based on the third preset category, obtain the similarity between the third classified audio and the third audio data, and obtain the similarity result; if the third preset category does not exist, obtain several fourth classified audios corresponding to the third preset category, obtain the similarity between all the fourth classified audios and the third audio data, and obtain the similarity result; obtain abnormal cattle based on the similarity result and the feeding time, track the abnormal cattle, obtain the abnormal detection result, and transmit the abnormal detection result to the user end.
[0033] Considering that cows have a short duration of estrus, this feature is related to the sex hormones secreted by the anterior pituitary gland, among which the follicle-stimulating hormone secreted by cows is the lowest among livestock, while the luteinizing hormone is the highest. Therefore, cows have a short duration of estrus and ovulate quickly. Due to this short estrus period, if you do not pay attention, it is very easy to miss the breeding opportunity. When cows are in estrus, they will have anorexia, mounting behavior, and change their calls. Their calls are usually low roars and long howls, which are different from the usual calls. These calls are mainly caused by changes in the hormone levels of cows. The present invention detects the estrus period through the above aspects, more accurately grasps its estrus cycle, and then accurately determines its breeding period, improves the breeding efficiency and the breeding performance of the herd.
[0034] Fence climbing behavior refers to the behavior of cows climbing on each other, mainly cows in estrus climbing on other cows.
[0035] Furthermore, the anomaly detection unit is also used for:
[0036] The abnormal genealogical relationship of the abnormal cattle is obtained, and an early warning result is obtained based on the abnormal genealogical relationship, and the early warning result is transmitted to the user end. Avoiding inbreeding and reducing genetic defects are more conducive to the growth and development of cattle, and improving the reproductive efficiency and reproductive performance of the herd.
[0037] Furthermore, a hanging rail is provided in the pasture, and the robot moves along the hanging rail, which makes it more convenient to detect ammonia concentration, collect cowshed images, and reduce damage to the robot caused by cows.
[0038] Furthermore, the cattle are equipped with an identification ring which displays the number, making identification, target location and tracking more convenient.
[0039] One or more technical solutions provided by the present invention have at least the following technical effects or advantages:
[0040] 1. Obtain a detection area based on the first detection area and the second detection area, obtain a second coordinate based on the detection area, move the robot to the second coordinate, control the ammonia detection equipment to detect the detection area, obtain the ammonia detection result, obtain a first concentration of ammonia based on the ammonia detection result, judge whether the first concentration is greater than the preset concentration, if so, obtain a second concentration of the ammonia remover based on the first concentration, and based on the second concentration, the robot controls the ammonia spraying equipment to spray the ammonia remover on the detection area; identify the number of cattle to obtain the distribution of cattle herds, obtain areas with many cattle herds, detect the ammonia concentration in the places where cattle herds gather, detect whether it affects the growth and development of cattle, detect areas without cattle herds but with a lot of feces and straw, and detect whether their living environment produces a large amount of ammonia. The two are combined for detection, the detection point is more flexible, more targeted, and the detection concentration is more accurate. The concentration of the ammonia remover is then configured according to the ammonia concentration, and the detection area is sprayed. Accurate spraying is performed on a small area to reduce the waste of agents.
[0041] 2. Obtain sick cows based on eye images, track sick cows, obtain eye detection results, transmit eye detection results to the user end, obtain cows with red eye disease according to pixel values, and track them to reduce the spread of red eye disease and ensure the normal growth of cows, as well as reflect cowshed environmental problems.
[0042] 3. Obtain the first number of the cattle in the detection area, determine whether there is a second number in the database that has a parent-child relationship with the first number, and if so, obtain the second audio data of the second number; based on the first preset category, obtain the first classification audio corresponding to the second audio data, and the playback device plays the first classification audio. After there are no cattle in the detection area, the ammonia spraying device sprays the detection area with ammonia remover, and by playing its audio, the cattle are moved away from the detection area and then the ammonia remover is sprayed to avoid the position of the cattle, thereby reducing the situation where the agent is sprayed on the cattle, which is not conducive to the growth of the cattle.
[0043] 4. Based on similar results and eating time, abnormal cattle are obtained, abnormal cattle are tracked, abnormal detection is obtained, and the abnormal detection results are transmitted to the user end; the estrus period of the cow is detected to more accurately grasp its estrus cycle, and then accurately determine its breeding period, thereby improving the reproductive efficiency and reproductive performance of the herd.
[0044] 5. Obtain the abnormal genealogical relationships of abnormal cattle, obtain early warning results based on the abnormal genealogical relationships, and transmit the early warning results to the user end; avoid inbreeding, reduce genetic defects, be more conducive to the growth and development of cattle, and improve the reproductive efficiency and reproductive performance of the herd. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of the present invention, and do not constitute a limitation on the embodiments of the present invention;
[0046] Figure 1 It is a system schematic diagram of the cattle farm intelligent inspection robot in the present invention;
[0047] Figure 2 It is a schematic diagram of the driving process of the driving unit in the cattle farm intelligent inspection robot of the present invention;
[0048] Figure 3 It is a schematic diagram of the detection process of the abnormality detection unit in the cattle farm intelligent inspection robot in the present invention. DETAILED DESCRIPTION
[0049] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.
[0050] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those within the scope of this description. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.
[0051] Example 1
[0052] refer to Figure 1 This embodiment provides a smart inspection robot for cattle farms, the robot comprising:
[0053] Collection unit: used to collect cattle farm data based on a collection device; in this embodiment, the collection device may be a camera;
[0054] Identification unit: used to obtain the cattle farm environment image and the cattle eye image based on the cattle farm data; specifically includes:
[0055] A first recognition unit: used to obtain a first image based on the cattle farm data, recognize the first image based on a recognition model to obtain a first recognition result, divide the cattle farm based on the first recognition result to obtain a plurality of sub-areas, and obtain the environmental image based on all the sub-areas; such as by recognizing cattle and calculating the number of cattle, and dividing them into a plurality of areas according to the number; in this embodiment, the recognition model can be a model trained by an image training set of labeled cattle, artificial intelligence, computer vision, deep learning and other technologies, which can automatically detect and recognize cattle, count the number, and recognize the facial features of cattle.
[0056] The second recognition unit is used to recognize the environment image based on the recognition model to obtain a cow image, recognize the cow image based on the recognition model to obtain a second recognition result; if the second recognition result is successful, obtain the eye image based on the second recognition result; if the second recognition result is unsuccessful, obtain a number of cows to be tested based on the second recognition result, obtain the images to be tested of the cows to be tested, recognize the images to be tested based on the recognition model to obtain facial features; obtain the feature quantity of the facial features, obtain a first coordinate based on the feature quantity and the facial features, the robot moves to the first coordinate, performs image acquisition based on the acquisition device, obtains a second image, and updates the cow image to the second image; in this embodiment, the facial features may include eyes, nose, mouth, horns, ears, etc.;
[0057] After the recognition model recognizes the cow, it detects whether there are two eyes in its facial area at the same time. If so, the recognition is successful, and its eye area is obtained to obtain the eye image. Otherwise, based on the recognized facial features, such as only recognizing the left half of the cow's facial area, the coordinates of the robot at this time are obtained. With the cow as the center of the circle, the robot is moved 90° clockwise to a position facing the cow's face.
[0058] Ammonia detection unit: used to obtain a detection area based on the environmental image, control the ammonia detection device to detect the detection area, and obtain an ammonia detection result; based on the detection result, control the ammonia spraying device to spray an ammonia remover on the cattle farm; specifically includes:
[0059] The first ammonia unit: used to obtain the number of cattle in each of the sub-areas based on the environmental image, and to sort all the sub-areas based on the number of cattle to obtain an area list; obtain the sub-areas within a preset ranking range in the area list to obtain a first detection area, such as obtaining the top 5 areas for detection; obtain the sub-area where the number of cattle is equal to 0, obtain the image to be tested, and obtain the feces area and the bedding area based on the image to be tested, such as using artificial intelligence, computer vision, deep learning and other technologies to identify feces and bedding, thereby obtaining the feces area and the bedding area, and respectively obtain the coordinates of the areas, thereby respectively obtaining the first area and the second area of the feces area and the bedding area, and obtaining the second detection area based on the first area and the second area, such as respectively selecting the areas with the largest areas to obtain the second detection area; obtain the detection area based on the first detection area and the second detection area;
[0060] A second ammonia unit: used to obtain a second coordinate based on the detection area, the robot moves to the second coordinate, controls the ammonia detection device to detect the detection area, and obtains an ammonia detection result; in this embodiment, the ammonia detection device may be an ammonia detector;
[0061] A third ammonia unit is used to obtain a first concentration of ammonia based on the ammonia detection result, determine whether the first concentration is greater than a preset concentration, and if so, obtain a second concentration of ammonia removal agent based on the first concentration. Based on the second concentration, the robot controls the ammonia spraying device to spray the ammonia removal agent on the detection area; in this embodiment, the ammonia spraying device may be a sprayer;
[0062] Eye detection unit: used to detect the cattle based on the eye image, obtain eye detection results, and transmit the eye detection results to the user end. Specifically used for:
[0063] The eye image is divided into several sub-images, and the pixel value of each sub-image is obtained. It is judged whether the pixel value is within a preset pixel range, such as whether it is within a red pixel range. If so, the sub-image is obtained to obtain a dangerous area, and the number of dangers in the dangerous area is obtained. It is judged whether the dangerous number is greater than a preset number. If it is greater than the preset number, sick cows are obtained based on the eye image, and the sick cows are tracked to obtain the eye detection result, and the eye detection result is transmitted to the user end.
[0064] Example 2
[0065] On the basis of the first embodiment, in this embodiment, the robot further includes a data unit, specifically including:
[0066] The first data unit is used to store the cattle number in the database;
[0067] The second data unit is used for collecting the first audio data of the cattle by an audio device based on the number, and storing the first audio data and the corresponding relationship between the number and the first audio data in the database; in this embodiment, the audio device may be a microphone,
[0068] The third data unit is used to divide the first audio data based on preset classifications to obtain a number of classified audios, such as dividing the audio into estrus audio of a cow, calling audio between parents and children, calling audio between non-parents and children, and danger warning audio, etc.; and store the corresponding relationship between the first audio data and all the classified audios, all the classified audios, and the category of each classified audio in the database;
[0069] The fourth data unit is used to obtain the genealogical relationship between the cattle, and store the genealogical relationship and the number corresponding to the genealogical relationship in the database.
[0070] Example 3
[0071] refer to Figure 2 , based on the above embodiment, in this embodiment, the ammonia detection unit further includes:
[0072] The driving unit is used to obtain the first number of the cattle in the detection area, determine whether there is a second number in the database that has a parent-child relationship with the first number, and if so, obtain the second audio data of the second number; based on the first preset category, obtain the first classification audio corresponding to the second audio data, and the playback device plays the first classification audio. After there are no cattle in the detection area, the ammonia spraying device sprays the ammonia removal agent on the detection area; in this embodiment, the playback device can be a speaker, etc.;
[0073] If not present, a number of second classification audios corresponding to the second preset category are obtained, and the playback device plays the second classification audios in sequence. After there are no cattle in the detection area, the ammonia spraying device sprays ammonia remover on the detection area.
[0074] If the cow is a cow, if there is a parent-child relationship, the audio data of the calf with which the parent-child relationship exists is obtained, and the calling audio is obtained from it, and the audio is played through a playback device to attract the cow to leave the detection area; if there is no parent-child relationship, the calling audio between non-parent-child relationships is obtained, and the audio is played through a playback device to attract the cow to leave the detection area.
[0075] Example 4
[0076] refer to Figure 3 , based on the above embodiment, in this embodiment, the robot further includes:
[0077] Anomaly detection unit: used to obtain the feeding time of the cattle based on the number and the preset time range;
[0078] Based on the recognition model, the image of the cattle is identified to obtain the cattle head area and the cattle area, and the third coordinates and fourth coordinates of the cattle head area and the cattle area are obtained respectively. Based on the third coordinates and the fourth coordinates, it is judged whether the cattle has fence climbing behavior, such as judging whether the cattle head is located above the cattle according to the coordinates. If so, the fence climbing cattle is obtained based on the third coordinates, and the third number and third audio data of the fence climbing cattle are obtained. The fourth audio data corresponding to the third number in the database is obtained, and it is judged whether the fourth audio data has a third preset category, such as judging whether there is audio data of estrus. If the third preset category exists, a third classified audio is obtained based on the third preset category, and the similarity between the third classified audio and the third audio data is obtained to obtain a similar result; if the third preset category does not exist, a number of fourth classified audios corresponding to the third preset category are obtained, and the similarity between all the fourth classified audios and the third audio data is obtained to obtain the similarity result; such as obtaining the similarity between audio data by audio feature extraction, machine learning methods, constructing perception models, and shift matching methods;
[0079] Abnormal cattle are obtained based on the similarity results and the eating time, the abnormal cattle are tracked, abnormal detection results are obtained, and the abnormal detection results are transmitted to the user end.
[0080] In this embodiment, the third audio data is also stored in a database and classified and stored according to preset categories.
[0081] Example 5
[0082] On the basis of the above embodiment, in this embodiment, the abnormality detection unit is further used to: obtain the abnormal genealogical relationship of the abnormal cattle, obtain an early warning result based on the abnormal genealogical relationship, and transmit the early warning result to the user end, such as transmitting the cattle numbers of close relatives to the user end.
[0083] Example 6
[0084] On the basis of the above embodiment, in this embodiment, a hanging rail is provided in the ranch, and the robot moves along the hanging rail.
[0085] Example 7
[0086] On the basis of the above embodiment, in this embodiment, the cattle wear an identification ring, and the identification ring displays the number.
[0087] Example 8
[0088] Based on the above embodiments, in this embodiment, the robot includes a collection device, an ammonia detection device, an ammonia spraying device, a controller, an audio device and a playback device; the controller is connected to the ammonia detection device, the ammonia spraying device, the audio device, the playback device, the user terminal and the identity ring, and the controller has a built-in recognition model and a data processing unit, and the data processing unit is connected to a database.
[0089] The controller is used to control the acquisition equipment to collect data, the ammonia detection equipment to detect ammonia concentration, the ammonia spraying equipment to spray ammonia removal agent, the audio equipment to collect cow calls, the playback equipment to play audio, the recognition model to identify targets and the data processing unit to process data.
[0090] The robot moves along the hanging rail. In this embodiment, the robot and the hanging rail can be connected by a telescopic rod to change the vertical distance, so that data can be collected more clearly and the ammonia remover can be sprayed more accurately.
[0091] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.
[0092] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.
Claims
1. The smart cruising robot for cattle farms is characterized by: The robot comprises: Collection unit: used to collect cattle farm data based on collection equipment; Recognition unit: used for obtaining an environment image of a cattle farm and an eye image of a cattle based on the cattle farm data; Ammonia detection unit: used to obtain a detection area based on the environmental image, control the ammonia detection device to detect the detection area, and obtain an ammonia detection result; based on the detection result, control the ammonia spraying device to spray an ammonia remover on the cattle farm; Eye detection unit: used for detecting the cattle based on the eye image, obtaining eye detection results, and transmitting the eye detection results to the user end; The identification unit specifically includes: A first recognition unit is used to obtain a first image based on the cattle farm data, recognize the first image based on a recognition model to obtain a first recognition result, divide the cattle farm based on the first recognition result to obtain a plurality of sub-areas, and obtain the environment image based on all the sub-areas; The second recognition unit is used to recognize the environmental image based on the recognition model to obtain a cow image, recognize the cow image based on the recognition model to obtain a second recognition result; if the second recognition result is successful, obtain the eye image based on the second recognition result; if the second recognition result is unsuccessful, obtain a number of cows to be tested based on the second recognition result, obtain the images to be tested of the cows to be tested, recognize the images to be tested based on the recognition model to obtain facial features; obtain the number of features of the facial features, obtain a first coordinate based on the number of features and the facial features, the robot moves to the first coordinate, performs image acquisition based on the acquisition device to obtain a second image, and updates the cow image to the second image; The ammonia detection unit specifically comprises: The first ammonia unit is used to obtain the number of cattle in each of the sub-areas based on the environmental image, and to sort all the sub-areas based on the number of cattle to obtain an area list; obtain the sub-areas within a preset ranking range in the area list to obtain a first detection area; obtain the sub-area where the number of cattle is equal to 0, obtain the image to be tested, obtain the feces area and the bedding area based on the image to be tested, obtain the first area and the second area of the feces area and the bedding area respectively, and obtain the second detection area based on the first area and the second area; obtain the detection area based on the first detection area and the second detection area; A second ammonia unit: used to obtain a second coordinate based on the detection area, the robot moves to the second coordinate, controls the ammonia detection device to detect the detection area, and obtains an ammonia detection result; A third ammonia unit: used for obtaining a first concentration of ammonia based on the ammonia detection result, determining whether the first concentration is greater than a preset concentration, and if so, obtaining a second concentration of ammonia removal agent based on the first concentration, and based on the second concentration, the robot controls the ammonia spraying device to spray the ammonia removal agent on the detection area; The eye detection unit is specifically used for: The eye image is divided into several sub-images, and the pixel value of each sub-image is obtained. It is judged whether the pixel value is within a preset pixel range. If so, the sub-image is obtained to obtain a dangerous area, and the number of dangers in the dangerous area is obtained. It is judged whether the dangerous number is greater than a preset number. If it is greater than the preset number, sick cows are obtained based on the eye image, and the sick cows are tracked to obtain the eye detection result, and the eye detection result is transmitted to the user end.
2. The cattle farm intelligent cruising robot according to claim 1, characterized in that: The robot also includes a data unit, which specifically includes: The first data unit is used to store the cattle number in the database; A second data unit is used for collecting the first audio data of the cattle by the audio device based on the number, and storing the first audio data and the corresponding relationship between the number and the first audio data in the database; The third data unit is used to divide the first audio data into several classified audios based on preset classifications, and store the corresponding relationship between the first audio data and all the classified audios, all the classified audios and the category of each classified audio in the database; The fourth data unit is used to obtain the genealogical relationship between the cattle, and store the genealogical relationship and the number corresponding to the genealogical relationship in the database.
3. The cattle farm intelligent cruising robot according to claim 2, characterized in that: The ammonia detection unit also includes: The driving unit is used to obtain the first number of the cattle in the detection area, determine whether there is a second number in the database that has a parent-child relationship with the first number, and if so, obtain the second audio data of the second number; based on the first preset category, obtain the first classification audio corresponding to the second audio data, and the playback device plays the first classification audio. After there are no cattle in the detection area, the ammonia spraying device sprays the ammonia remover on the detection area; If not present, a number of second classification audios corresponding to the second preset category are obtained, and the playback device plays the second classification audios in sequence. After there are no cattle in the detection area, the ammonia spraying device sprays ammonia remover on the detection area.
4. The cattle farm intelligent cruising robot according to claim 3, characterized in that: The robot also includes: Anomaly detection unit: used to obtain the feeding time of the cattle based on the number and the preset time range; Based on the recognition model, the image of the cattle is recognized to obtain the cattle head area and the cattle area, and the third coordinates and fourth coordinates of the cattle head area and the cattle area are obtained respectively. Based on the third coordinates and the fourth coordinates, it is judged whether the cattle has fence climbing behavior. If so, the fence climbing cattle is obtained based on the third coordinates, and the third number and third audio data of the fence climbing cattle are obtained. The fourth audio data corresponding to the third number in the database is obtained, and it is judged whether the fourth audio data has a third preset category. If the third preset category exists, a third classified audio is obtained based on the third preset category, and the similarity between the third classified audio and the third audio data is obtained to obtain a similarity result; if the third preset category does not exist, a number of fourth classified audios corresponding to the third preset category are obtained, and the similarity between all the fourth classified audios and the third audio data is obtained to obtain the similarity result; Abnormal cattle are obtained based on the similarity results and the eating time, the abnormal cattle are tracked, abnormal detection results are obtained, and the abnormal detection results are transmitted to the user end.
5. The cattle farm intelligent cruising robot according to claim 4, characterized in that: The anomaly detection unit is also used for: The abnormal pedigree relationship of the abnormal cattle is obtained, an early warning result is obtained based on the abnormal pedigree relationship, and the early warning result is transmitted to the user end.
6. The cattle farm intelligent cruising robot according to claim 1, characterized in that: A hanging rail is arranged in the ranch, and the robot moves along the hanging rail.
7. The cattle farm intelligent cruising robot according to claim 2, characterized in that: The cattle wear an identification ring which displays the number.
