Machine vision-based intelligent autonomous navigation positioning system with body
By using a machine vision-based embodied intelligent autonomous navigation and positioning system, combined with obstacle avoidance route status analysis and status evaluation modules, the obstacle avoidance route is dynamically adjusted, solving the problem of single obstacle avoidance route planning in existing technologies. This achieves highly adaptable obstacle avoidance route planning, meeting the transportation needs of different goods.
Patent Information
- Application Number
- CN202511268855.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-06
- Publication Date
- 2026-01-16
AI Technical Summary
The existing obstacle avoidance route planning methods are too simplistic and cannot adapt to the differences in the types of goods transported, thus failing to meet actual transportation needs.
The system utilizes a transportation route navigation planning module, a visual capture module, an obstacle avoidance route status analysis module, a status score evaluation module, and a preferred route selection module to formulate and optimize obstacle avoidance routes. It also dynamically adjusts obstacle avoidance routes based on the obstacle avoidance status of transportation equipment and the type of goods to meet the needs of different goods.
It enables adaptive adjustments to obstacle avoidance routes, improves the adaptability of obstacle avoidance routes, meets the transportation needs of different types of goods, and improves transportation efficiency and safety.
Smart Images
Figure CN121346788A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the field of visual capture technology, in particular, to a somatic intelligent autonomous navigation positioning system based on machine vision. BACKGROUND
[0002] Somatic intelligence is a frontier direction in the field of artificial intelligence, and its core is to realize the emergence and evolution of intelligence through physical entity and environment interaction. This concept emphasizes the decisive role of the coordinated evolution of body shape, perception system and action ability in the formation of intelligence.
[0003] Somatic intelligence is also widely used in the logistics industry, such as intelligent transport vehicles, which realize automatic navigation and route planning functions for cargo classification and transportation in industrial parks.
[0004] In the process of making route planning for transport vehicles, due to the large number of vehicles in the park, the inherent route is often occupied by other vehicles. The existing processing method mainly captures obstacles on the transport route in advance through machine vision, and then plans an obstacle avoidance route. The planning conditions of the obstacle avoidance route are often fixed, such as reducing the distance, but due to the different types of transported goods, there are differences in the matching planning conditions, and these planning conditions need to be actually passed through the obstacle avoidance route to obtain, such as the flatness of the obstacle avoidance route. The existing obstacle avoidance route planning method is too single and cannot adapt to the type of transported goods.
[0005] In order to solve the above problems, it is urgent to provide a somatic intelligent autonomous navigation positioning system based on machine vision which can adaptively simulate. SUMMARY
[0006] The purpose of the present application is to provide a somatic intelligent autonomous navigation positioning system based on machine vision, which formulates an obstacle avoidance route action sequence through an obstacle avoidance route state analysis module, assigns a corresponding obstacle avoidance route according to the moving order of the transport equipment, collects information using the transport equipment actually passing through the obstacle avoidance route, collects the obstacle avoidance state of the transport equipment corresponding to each obstacle avoidance route action, and combines the obstacle avoidance state of the obstacle avoidance route with the state evaluation module to evaluate the state score corresponding to the obstacle avoidance route. The obstacle avoidance route is selected by the state score to adapt to the type of transported goods of the actual transport equipment, so as to solve the problems raised in the background art, that is:
[0007] The existing obstacle avoidance route planning method is too single and cannot adapt to the type of transported goods.
[0008] To achieve the above purpose, a somatic intelligent autonomous navigation positioning system based on machine vision is provided, which includes a transport route navigation planning module, a visual capture module, an obstacle avoidance route state analysis module, a state score evaluation module and an optimal route selection module.
[0009] The initial route is planned by the transportation route navigation planning module, and the corresponding obstacle avoidance route is formulated in cooperation with the obstacle information captured by the visual capture module;
[0010] In the actual transportation process, the obstacle avoidance routes are sequentially distributed to the corresponding transportation devices for actual route state sensing, and the obstacle avoidance states of the transportation devices corresponding to the actions of the obstacle avoidance routes are collected by the obstacle avoidance route state analysis module, wherein the obstacle avoidance states reflect the current road condition state of the obstacle avoidance route;
[0011] Meanwhile, the state score is introduced for obstacle avoidance state evaluation, and the state score corresponding to the obstacle avoidance route is evaluated in cooperation with the state score evaluation module and the obstacle avoidance state of the obstacle avoidance route. The state score of each obstacle avoidance route is evaluated in terms of individual obstacle avoidance state and overall state, so as to cooperate with the subsequent cargo type division. The obstacle avoidance states of the obstacle avoidance routes adapted to different cargo types are different, and therefore the obstacle avoidance route adapted to the transportation device is selected by the optimal route selection module in combination with the actual value of the state score, and the adaptive obstacle avoidance route is selected in combination with the state score evaluation of the corresponding obstacle avoidance route.
[0012] Compared with the prior art, the present application has the following advantages:
[0013] In the embodiment of the present application, the obstacle avoidance route action sequence is formulated by the obstacle avoidance route state analysis module, the corresponding obstacle avoidance route is distributed to the transportation device according to the moving order, the information is collected by the transportation device actually passing through the obstacle avoidance route, the obstacle avoidance state of the transportation device corresponding to the action of each obstacle avoidance route is collected, and the state score corresponding to the obstacle avoidance route is evaluated in cooperation with the state score evaluation module and the obstacle avoidance state of the obstacle avoidance route. The obstacle avoidance route is selected by the state score to adapt to the transportation cargo type of the actual transportation device, and the obstacle avoidance route meeting the actual demand of the transportation cargo of different types is matched, so that the adaptive effect of the obstacle avoidance route is improved. BRIEF DESCRIPTION OF DRAWINGS
[0014] Figure 1 The overall structure diagram of the present application is shown in the figure.
[0015] The meanings of the various numbers in the figure are as follows:
[0016] 10, transportation route navigation planning module;
[0017] 20, visual capture module;
[0018] 30, obstacle avoidance route state analysis module;
[0019] 40, state score evaluation module;
[0020] 50, optimal route selection module. DETAILED DESCRIPTION
[0021] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.
[0022] Please refer to Figure 1 As shown in the figure, a machine vision-based embodied intelligent autonomous navigation positioning system is provided, including a transportation route navigation planning module 10, a vision capture module 20, an obstacle avoidance route state analysis module 30, a state score evaluation module 40, and an optimal route selection module 50.
[0023] The transportation route navigation planning module 10 is configured with a GPS navigation device, which can locate the position of the transportation device in real time and plan an initial route according to the positions of the transportation starting point and the destination.
[0024] The vision capture module 20 is used to capture route images during the movement of the transportation device, capture obstacles affecting the movement of the transportation device within the route range, and calculate the distance between the obstacles and the transportation device in the parking state and the shielding range of the initial route by using a laser scanning ranging method.
[0025] The transportation route navigation planning module 10 responds to the obstacle capture information, cooperates with the GPS navigation device to obtain the position coordinates of the current obstacle, and formulates multiple obstacle avoidance routes.
[0026] The obstacle avoidance route state analysis module 30 formulates an action sequence of the obstacle avoidance route, allocates the corresponding obstacle avoidance route according to the movement order of the transportation device, and collects the corresponding obstacle avoidance state of the transportation device for each action of the obstacle avoidance route.
[0027] The state score evaluation module 40 combines the obstacle avoidance state of the obstacle avoidance route to evaluate the state score corresponding to the obstacle avoidance route.
[0028] The optimal route selection module 50 combines the actual value of the state score to select the obstacle avoidance route suitable for the transportation device.
[0029] The specific content is as follows:
[0030] Firstly, since the starting point and the destination of the transportation device are different, the position of the transportation device needs to be located in real time by the transportation route navigation planning module 10, and an initial route is planned according to the positions of the transportation starting point and the destination, that is, the configured GPS navigation system is used for positioning processing to obtain the geographic coordinates of the transportation starting point and the destination, the corresponding initial route is planned according to the distribution of the industrial park map, and the transportation device acts according to the set initial route.
[0031] In the process of action, the route image in the moving process of the transport equipment is captured by using the visual capture module 20, the obstacles affecting the movement of the transport equipment in the route range are captured, the distance between the obstacles and the transport equipment in the parking state and the shielding range of the initial route are calculated by using the laser scanning ranging method, and the specific method is as follows:
[0032] Firstly, the offset angle of the laser scanning ranging method is adjusted, the distances corresponding to different obstacle reflection points are captured, and the distance formula is used, wherein s is the distance between the obstacle reflection point and the transport equipment, s1=s*cosθ when the laser scanning device has an angle offset, wherein θ is the offset angle generated in the measurement process of the laser scanning device, v0 is the propagation speed of light in air, t is the time spent for laser emission to the obstacle and reflection from the obstacle to the laser device, the distances of multiple reflection points are obtained, the final average distance is calculated as the distance between the current transport equipment parking position and the obstacle, and for the shielding range, the laser scanning device is uniformly deflected along the horizontal line, that is, a fixed angle is rotated each time, the distances of the corresponding reflection points are obtained in turn, the distances of two adjacent reflection points are obtained according to the Pythagorean theorem, and finally the distances of the two edge reflection points of the obstacle in the horizontal direction are calculated to obtain the transverse length of the obstacle. Of course, due to the different shapes and structures of the obstacles, different heights of the transverse length need to be captured to obtain the shielding range of the current obstacle in the actual capture process. The highest transverse length is obtained as the shielding range of the current obstacle. Since it is not considered to pass from above or below the obstacle, the height of the obstacle is not considered.
[0033] After the obstacle is identified, the obstacle needs to be avoided, and the obstacle avoidance route planning is performed. The transport route navigation planning module 10 responds to the obstacle capture information, cooperates with the GPS navigation device to obtain the position coordinates of the current obstacle, and formulates multiple obstacle avoidance routes. At this time, the current route position of the obstacle, the transverse length of the obstacle and the transverse length of the transport equipment need to be combined and determined. When the transverse length of the current transport equipment is lower than the distance reserved between the obstacle and the road, the route can still be used as the obstacle avoidance route of the transport equipment at this time. At this time, the position of the obstacle on the route is combined, and the laser scanning device guides the transport equipment to pass through the distance. Conversely, when the transverse length of the current transport equipment is not lower than the distance reserved between the obstacle and the road, it indicates that the current initial route cannot pass. Of course, for the remaining obstacle avoidance routes, the industrial park route distribution map and the existing map navigation software are combined to plan, and the remaining obstacle avoidance routes are planned;
[0034] The obstacle avoidance route state analysis module 30 formulates an obstacle avoidance route action sequence. In the present scheme, when the current initial route cannot be passed, the distance of each obstacle avoidance route is used to plan the action sequence, that is, the shorter the distance, the earlier the action sequence. However, in actual transportation, the single transportation capacity of the transportation device often cannot meet the actual transportation demand, so multiple transportation devices need to be used for circular transportation. Therefore, the obstacle avoidance state of each obstacle avoidance route needs to be fed back by actual transportation, so the corresponding obstacle avoidance route needs to be allocated according to the moving order of the transportation device, that is, the transportation device that starts first uses the obstacle avoidance route with the shortest distance. It is worth noting that even if the obstacle avoidance route is still on the current initial route, since the route has changed at this time, the corresponding obstacle avoidance route needs to be started, that is, as long as the transportation device deviates from the original initial route, the corresponding avoidance route needs to be planned. When each obstacle avoidance route passes through the transportation device, since the transportation device passes through the corresponding obstacle avoidance route, the internal sensing device of the transportation device can be used to collect the obstacle avoidance state of the transportation device corresponding to the action of each obstacle avoidance route, wherein the obstacle avoidance state includes vibration frequency, vehicle inclination and moving speed, for example, the corresponding sensor is monitored, but it is not limited to the above obstacle avoidance state. In actual use, it is determined according to the needs of the goods in the transportation device, for example, when the goods in the transportation device are liquid, if there is a road unevenness in the corresponding obstacle avoidance route at this time, it will cause the transportation device to vibrate, which is easy to cause liquid overflow, so the type of goods needs to be adapted to select the obstacle avoidance route according to the obstacle avoidance state.
[0035] In the actual selection process, the state evaluation module 40 needs to evaluate the state score of the obstacle avoidance route in combination with the obstacle avoidance state of the obstacle avoidance route. The specific content is as follows:
[0036] Firstly, the initial integral In of each obstacle avoidance state is defined, and the actual value of each obstacle avoidance state is obtained by combining the configured sensor, wherein V f is the vibration frequency value monitored when the transportation device passes through the obstacle avoidance route, In f is the initial integral of the vibration frequency, is the state score of the vibration frequency, V n is the vehicle inclination value monitored when the transportation device passes through the obstacle avoidance route, In n is the initial integral of the vehicle inclination, is the state score of the vehicle inclination, m p is the moving speed value monitored when the transportation device passes through the obstacle avoidance route, In p is the initial integral of the moving speed, m p x In p is the state score of the moving speed.
[0037] Finally, the preferred route selection module 50 combines the actual state value with the state score to select an obstacle avoidance route suitable for the transportation equipment. Since the transportation equipment transports different goods, the types of obstacle avoidance routes it adapts to also differ. Therefore, when matching the preferred route, the route inclination degree needs to be analyzed in advance according to the type of goods transported by the transportation equipment. Different types of goods are divided into vibration frequency type, vehicle inclination type, moving speed type, and conventional type. These division methods are determined manually, for example, for fresh produce, both moving speed and road conditions need to be considered, so vibration frequency and vehicle inclination are considered. At this time, this type of goods will be divided into the conventional type, that is, all obstacle avoidance states need to be considered, and the final state score is calculated as the evaluation standard. For goods of a single type, the corresponding obstacle avoidance state score is used for division. The obstacle avoidance route is divided according to the size of the state score corresponding to each obstacle avoidance route. The higher the state score, the higher the preference degree of the current obstacle avoidance route, and the more suitable it is as the preferred route for the current goods.
[0038] The obstacle avoidance route state analysis module 30 of the present application formulates an obstacle avoidance route action sequence, assigns the corresponding obstacle avoidance route according to the transportation equipment moving order, collects information using the actual obstacle avoidance route of the transportation equipment, collects the corresponding obstacle avoidance state of the transportation equipment for each obstacle avoidance route action, and combines the obstacle avoidance state of the obstacle avoidance route with the state score evaluation module 40 to evaluate the state score corresponding to the obstacle avoidance route. The obstacle avoidance route is selected by the state score to adapt to the type of goods transported by the actual transportation equipment, match the obstacle avoidance route that meets the actual needs of different types of transportation goods, and improve the adaptation effect of the obstacle avoidance route.
[0039] The basic principles, main features and advantages of the present application are shown and described above. Those skilled in the art should understand that the present application is not limited by the above examples, and the above examples and descriptions in the specification are only preferred examples of the present application and do not limit the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.
Claims
1. A machine vision based embodied intelligent autonomous navigation and localization system, characterized in that: The application relates to a transport route navigation planning system, which comprises a transport route navigation planning module (10), a visual capture module (20), an obstacle avoidance route state analysis module (30), a state score evaluation module (40) and a preferred route selection module (50). The transport route navigation planning module (10) is provided with a GPS navigation device, which can locate the position of the transport device in real time and plan an initial route according to the positions of the starting point and the destination of the transport. The visual capture module (20) is used for capturing the route images during the movement of the transport device, capturing the obstacles affecting the movement of the transport device in the route range, calculating the distance between the obstacles and the transport device in the parking state and the shielding range of the initial route by using a laser scanning ranging method. The transport route navigation planning module (10) responds to the obstacle capture information, obtains the position coordinates of the current obstacles by cooperating with the GPS navigation device, and plans multiple obstacle avoidance routes. The obstacle avoidance route state analysis module (30) plans the action sequence of the obstacle avoidance routes, assigns the corresponding obstacle avoidance routes according to the movement sequence of the transport device, and collects the corresponding obstacle avoidance states of the transport device in the action of each obstacle avoidance route. The state score evaluation module (40) evaluates the state score corresponding to the obstacle avoidance route in combination with the obstacle avoidance state of the obstacle avoidance route. The preferred route selection module (50) selects the obstacle avoidance route suitable for the transport device in combination with the actual value of the state score.
2. The machine vision based embodied intelligent autonomous navigation and positioning system according to claim 1, wherein: The method for calculating the distance between the obstacles and the transport device in the parking state in the visual capture module (20) comprises the following steps: S201, adjusting the offset angle of the laser scanning ranging method to capture the distances corresponding to different obstacle reflection points; S202, utilize the distance formula Wherein s is the distance between the obstacle reflection point and the transportation equipment, when the laser scanning equipment has an angle offset, the corresponding s1=s*cosθ, wherein θ is the generated angle of deviation in the measurement process of the laser scanning equipment, v0is the propagation speed of light in air, t is the time spent for laser emission to the obstacle and reflection from the obstacle to the laser equipment, and the distances of multiple reflection points are obtained; S203, calculating the final average distance as the distance between the current transport device parking position and the obstacle.
3. The machine vision based embodied intelligent autonomous navigation and positioning system according to claim 2, wherein: The method for calculating the shielding range of the initial route by the obstacles in the visual capture module (20) comprises the following steps: S204, uniformly deflecting along the horizontal line by using the laser scanning device; S205, sequentially obtaining the distances of the corresponding reflection points and obtaining the distances of two adjacent reflection points according to the Pythagorean theorem; S206, calculating the distances of the two edge reflection points of the obstacle in the horizontal direction to obtain the transverse length of the obstacle.
4. The machine vision based embodied intelligent autonomous navigation and positioning system according to claim 1, wherein: The method for planning the obstacle avoidance route in the transport route navigation planning module (10) comprises the following steps: S101, combining the route position of the current obstacle, the transverse length of the obstacle and the transverse length of the transport device, and making a judgment; When the transverse length of the current transport device is lower than the distance reserved between the obstacle and the road, the current initial route is used as the obstacle avoidance route of the transport device, and the laser scanning device is used to guide the transport device to pass through the distance in combination with the route position of the obstacle; When the transverse length of the current transport device is not lower than the distance reserved between the obstacle and the road, it is indicated that the current initial route cannot pass through; S102, planning the remaining obstacle avoidance routes in combination with the industrial park route distribution map and the existing map navigation software, the parking position of the current transport device and the destination position.
5. The machine vision based embodied intelligent autonomous navigation and positioning system according to claim 1, wherein: 6. The machine vision based embodied intelligent autonomous navigation and positioning system according to claim 5, wherein: The obstacle avoidance state collected in the obstacle avoidance route state analysis module (30) includes vibration frequency, vehicle inclination and moving speed.
7. The machine vision based embodied intelligent autonomous navigation and positioning system according to claim 1, wherein: The method for the state score evaluation module (40) to evaluate the state score corresponding to the obstacle avoidance route includes the following steps: S401, defining the initial integral In of each obstacle avoidance state; S402, the actual values of each obstacle avoidance state are obtained in combination with the configured sensors, and the actual values are calculated where V f is the number of vibration frequencies monitored by the transport equipment as it traverses the obstacle-avoiding route, In f is the initial integration of the vibration frequencies, is the state fraction of vibration frequencies; V n for the vehicle inclination value monitored by the transport device when passing the obstacle avoidance route, In n for the initial integration of the vehicle inclination, for the state division of the vehicle inclination; m p The value of the moving speed monitored when the transport equipment passes through the obstacle avoidance route, In p For the initial integral of the moving speed, m p ×In p The state is determined by the speed of movement.
8. The machine vision based embodied intelligent autonomous navigation and positioning system according to claim 1, wherein: The method for the preferred route selection module (50) to select the obstacle avoidance route suitable for the transportation equipment includes the following steps: S501, performing route tendency analysis according to the type of goods transported by the transportation equipment, and dividing different types of goods into vibration frequency type, vehicle inclination type, moving speed type and conventional type; S502, for the type of goods divided into a single type, using the corresponding obstacle avoidance state score for division; S503, for the type of goods divided into multiple types, using the obstacle avoidance state score of the corresponding type for division; S504, for the type of goods divided into all types, using the overall state score for division.