An intelligent agricultural visual control system based on finger touch self-power generation standardized light pulse coding

Abstract

The application discloses a kind of intelligent agricultural visual control systems based on finger touch self-generating standard light pulse coding, including self-generating touch control panel, visual acquisition module, light coding identification module, instruction analysis module and agricultural equipment safety control module.Finger contact and leave friction layer instant friction nanometer power generation unit output pulse electric energy by operator;The branch rectifies, energy storage and threshold trigger to the electric energy, and drive LED array output and visual acquisition frame rate matching standardized LED light pulse.LED array generates space-time composite light coding signal.Light coding identification module determines target identification area, extracts LED brightness time sequence, and judges whether signal is effective or not.Instruction analysis module resolves light coding signal into agricultural equipment control instruction, and agricultural equipment safety control module executes or prohibits corresponding control instruction.The application does not need touch end battery, and is suitable for low false trigger visual control in complex agricultural light environment.

Description

Technical Field

[0001] This invention relates to the fields of intelligent agriculture, human-computer interaction, self-driven sensing, triboelectric nanogenerator, LED light coding, machine vision recognition and agricultural equipment control technology, and in particular to an intelligent agricultural vision control system based on finger touch self-generated standardized light pulse coding. Background Technology

[0002] In smart agriculture scenarios, actuators such as water pumps, solenoid valves, fans, supplemental lighting, spraying equipment, fertilizing equipment, and rolling shutters typically require start-stop control based on manual commands or environmental parameters. Existing control methods mostly employ wired buttons, wireless remote controls, mobile terminals, or IoT gateways. While these methods enable equipment control, touch-screen or remote control terminals usually rely on batteries or external power supplies. In high-humidity, misty, dusty, and high-light environments such as greenhouses, fields, and orchards, these methods suffer from high power supply maintenance costs, easy interface corrosion, complex wiring, and insufficient wireless communication stability.

[0003] Triboelectric nanogenerators can convert low-frequency mechanical energy from finger contact, separation, pressing, or sliding into electrical energy, making them suitable for building battery-free touch input devices. However, the electrical energy generated by finger touch is transient, random, and fluctuates in amplitude. Directly driving LEDs can easily lead to problems such as excessively short light emission time, unstable brightness, and missed detections due to discrete frame sampling by the camera. Furthermore, agricultural environments are subject to interference from sunlight reflection, flickering supplemental lighting, water mist reflection, greenhouse film reflection, mud and dust contamination, and changes in camera installation angle. Ordinary single-frame recognition or fixed threshold recognition can easily trigger agricultural equipment falsely. Therefore, an intelligent agricultural vision control system is needed that can convert the unstable transient electrical energy generated by finger touch into standardized light pulses suitable for stable camera acquisition, and combine optical encoding verification, visual anti-interference recognition, and agricultural safety interlock control. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent agricultural vision control system based on finger-touch self-generated standardized light pulse coding, in order to solve the problems of existing agricultural control terminals relying on batteries or external power supply, transient instability of finger-touch power generation signals, easy missed detection of camera frame sampling, easy misidentification of light signals in complex agricultural lighting environments, and high risk of malfunction of agricultural execution equipment.

[0005] To achieve the above objectives, the present invention adopts the following technical solution.

[0006] A smart agricultural vision control system based on finger-touch self-generated standardized optical pulse coding includes a self-generated touch control panel, a vision acquisition module, an optical coding recognition module, an instruction parsing module, and an agricultural equipment safety control module. The self-generated touch control panel includes multiple touch-sensing areas, multiple single-electrode triboelectric nanogenerators, an energy shaping and threshold-triggered delayed light-emitting branch, an LED array, and positioning markers. Each touch-sensing area corresponds to at least one single-electrode triboelectric nanogenerator, which includes a triboelectric layer disposed on the surface of the touch-sensing area and an electrode layer disposed below the triboelectric layer. When the operator's finger or touch contact object touches and leaves the triboelectric layer, the finger or touch contact object contacts and electrifies the triboelectric layer, generating induced charge changes in the electrode layer and outputting transient pulse electrical energy.

[0007] Furthermore, the energy shaping and threshold-triggered delayed light emission branch is electrically connected to the electrode layer and the LED array, and is used for rectifying, storing, limiting, threshold-triggered, and discharging transient pulse energy. When the energy storage voltage reaches a preset trigger threshold, the energy shaping and threshold-triggered delayed light emission branch drives at least one LED light-emitting unit in the LED array to output a standardized LED light pulse; when the energy storage voltage does not reach the preset trigger threshold, it prohibits the output of incomplete or low-brightness light pulses. The standardized LED light pulse includes at least the output start condition defined by the preset trigger threshold, a duration matching the frame rate of the visual acquisition module, and a detectable brightness that meets the recognition threshold, thereby reducing the risk of missed detection and misidentification caused by weak touch, wet hand touch, gloved touch, or short-term touch.

[0008] Furthermore, the LED array is used to generate a spatial-temporal composite optical coding signal. The spatial-temporal composite optical coding signal includes a device coding segment, an action coding segment, and a verification coding segment. The device coding segment is determined by the spatial position of the LED light-emitting units and is used to indicate the type of agricultural execution equipment; the action coding segment is determined by the number of flashes, pulse width, light emission duration, or flash interval of the LED light-emitting units and is used to indicate start, stop, gear switching, mode switching, or emergency stop actions; the verification coding segment is determined by the start pulse, end pulse, repeat pulse, or confirmation pulse of at least one verification LED and is used to determine whether the control command is complete and valid.

[0009] Furthermore, the visual acquisition module is used to acquire images or video sequences containing the LED array. The optical encoding recognition module determines the target recognition area of ​​the LED array based on the positioning marks set on the self-generating touch control panel, and extracts the brightness time series of each LED light-emitting unit within the target recognition area. The optical encoding recognition module determines whether the spatial-temporal composite optical encoding signal is valid based on spatial location, dynamic brightness threshold, number of consecutive frames, time series rules, and verification encoding segments. Preferably, the optical encoding recognition module can map the LED array area to a preset standard coordinate system based on the positioning marks, and perform recognition based on the difference between the brightness of the LED points and the local background brightness, so as to reduce misidentification caused by sunlight reflection, water mist reflection, supplementary lighting flicker, greenhouse film reflection, and camera installation deviation.

[0010] Furthermore, the instruction parsing module is used to parse spatial-temporal composite optical coded signals that meet preset rules into agricultural equipment control instructions. The agricultural equipment safety control module is used to control the operation of the corresponding agricultural equipment based on the agricultural equipment control instructions, the current operating status of the agricultural execution equipment, mutually exclusive instruction relationships, control instruction priorities, and preset safety conditions; when there is equipment malfunction, mutually exclusive instructions, unmet environmental conditions, or unmet safety conditions, the execution of the corresponding agricultural equipment control instructions is prohibited. The agricultural execution equipment includes, but is not limited to, water pumps, solenoid valves, fans, supplemental lighting, spraying equipment, fertilizer pumps, nutrient solution circulation pumps, rolling shutter equipment, wet curtains, greenhouse window opening mechanisms, or agricultural robots.

[0011] Compared with the prior art, the present invention has at least the following beneficial effects: First, this invention uses a finger-contact-separated single-electrode triboelectric nanogenerator unit in conjunction with a threshold-triggered delayed light-emitting branch to convert unstable transient touch power into standardized LED light pulses that match the camera frame rate, thereby reducing the risk of missed detection and misidentification caused by weak touch, wet hands, gloved touch, or short-term touch.

[0012] Second, the present invention constructs a spatial-temporal composite optical code by means of a device coding segment, an action coding segment, and a verification coding segment, which enables the expansion of control commands for various types of agricultural equipment under the conditions of limited LED quantity and low energy input, and improves the command integrity verification capability.

[0013] Third, this invention achieves stable visual recognition in complex agricultural lighting environments through positioning markers, target recognition areas, brightness time series, dynamic thresholds, and multi-frame verification, and reduces the risk of malfunctions in irrigation, supplemental lighting, spraying, and ventilation equipment through safety interlock control of agricultural equipment. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the overall structure of the intelligent agricultural vision control system of the present invention; Figure 2 This is a schematic diagram of the self-generating touch control panel of the present invention; Figure 3 This is a schematic diagram of the structure of the single-electrode triboelectric nanogenerator unit and the energy shaping and threshold-triggered delayed light emission branch of the present invention; Figure 4 This is a schematic diagram of the LED array spatial-temporal composite optical coding, positioning mark and brightness recognition of the present invention; Figure 5 This is a flowchart illustrating the optical encoding recognition, multi-frame verification, and agricultural equipment safety control process of this invention.

[0015] Explanation of reference numerals in the attached figures 1. Self-generating touch control panel; 11. Touch sensing area; 12. Single-electrode triboelectric nano-power generation unit; 121. Triboelectric layer; 122. Electrode layer; 13. Energy shaping and threshold-triggered delayed light emission branch; 131. Rectifier circuit; 132. Energy storage capacitor; 133. Voltage threshold triggering unit; 134. Discharge control unit; 135. Low-power LED driver unit; 14. LED array; 141. LED light emission unit; 142. Verification LED; 15. Positioning mark; 2. Visual acquisition module; 3. Optical encoding recognition module; 4. Command parsing module; 5. Agricultural equipment safety control module; 6. Agricultural execution equipment; 7. Point window; 8. Local background window. Detailed Implementation

[0016] The technical solution of the present invention will be further described below with reference to the accompanying drawings. It should be understood that the following embodiments are only used to illustrate the present invention and are not intended to limit the scope of protection of the present invention. Without departing from the concept of the present invention, any equivalent substitutions or conventional modifications made by those skilled in the art regarding the number of touch sensing areas, triboelectric materials, LED array forms, encoding rules, identification methods, or types of agricultural execution devices should be included within the scope of protection of the present invention.

[0017] like Figure 1As shown, the intelligent agricultural vision control system based on finger-touch self-generated standardized light pulse coding provided by the present invention includes a self-generated touch control panel 1, a vision acquisition module 2, a light coding recognition module 3, an instruction parsing module 4, and an agricultural equipment safety control module 5. The self-generated touch control panel 1 is used to convert the operator's finger touch actions into standardized LED light pulses and form light-coded signals; the vision acquisition module 2 is used to acquire images or video sequences containing an LED array 14; the light coding recognition module 3 is used to perform target area positioning, brightness time series extraction, and multi-frame verification on the LED light-coded signals; the instruction parsing module 4 is used to parse the valid light-coded signals into agricultural equipment control instructions; and the agricultural equipment safety control module 5 is used to control the operation of the agricultural execution equipment 6 according to the agricultural equipment control instructions and safety conditions.

[0018] In this embodiment, the self-generated touch control panel 1 can be installed at the greenhouse entrance, outside the agricultural control cabinet, on a field pillar, on a plant factory operating platform, in an agricultural robot docking area, or in other locations that are easily accessible for manual touch control and can be observed by the visual acquisition module 2. The visual acquisition module 2 can be a greenhouse monitoring camera, an industrial camera, a network camera, a mobile terminal camera, an infrared camera, or an agricultural robot camera. The agricultural execution equipment 6 includes, but is not limited to, water pumps, solenoid valves, fans, supplemental lighting, spraying equipment, fertilizer pumps, nutrient solution circulation pumps, rolling shutter equipment, wet curtains, greenhouse window opening mechanisms, or agricultural robots. The visual acquisition module 2 can be set up independently or can reuse an existing crop monitoring camera in the greenhouse. When the positioning mark 15 is detected in the image and an LED light pulse that meets the rules is identified, the system enters the light encoding recognition mode; when no valid light encoding signal is detected, the visual acquisition module 2 can continue to perform crop growth monitoring, pest and disease observation, or environmental inspection tasks.

[0019] like Figure 2 As shown, the self-generating touch control panel 1 includes multiple touch sensing areas 11, multiple single-electrode triboelectric nanogenerators 12, energy shaping and threshold-triggered delayed light-emitting branches 13, an LED array 14, and positioning markers 15. The multiple touch sensing areas 11 can be arranged in a matrix or according to agricultural equipment control function zones. For example, the touch sensing areas 11 correspond to irrigation control, supplemental lighting control, ventilation control, spraying control, fertilization control, and emergency stop control, respectively. Each touch sensing area 11 can be equipped with text, graphics, raised textures, color codes, or blind-touch recognition structures to allow operators to identify the corresponding function under conditions of humidity, strong light, or when wearing gloves.

[0020] like Figure 2 and Figure 3As shown, each touch sensing area 11 corresponds to at least one single-electrode triboelectric nanogenerator unit 12. The single-electrode triboelectric nanogenerator unit 12 includes a triboelectric layer 121 and an electrode layer 122. The triboelectric layer 121 is disposed on the surface of the touch sensing area 11 for direct contact with the operator's finger or touch contact body; the electrode layer 122 is disposed below the triboelectric layer 121 and is electrically connected to the energy shaping and threshold-triggered delayed light emission branch 13. When the operator's finger or touch contact body contacts the triboelectric layer 121, a contact charge is generated between them; when the operator's finger or touch contact body leaves the triboelectric layer 121, the change in the relative distance between the finger or touch contact body and the electrode layer 122 causes a redistribution of the induced charge in the electrode layer 122, thereby outputting transient pulse electrical energy.

[0021] As a preferred technical solution, the triboelectric layer 121 can be made of polytetrafluoroethylene, fluorinated ethylene propylene copolymer, polydimethylsiloxane, polyimide, polyvinylidene fluoride, polyester film, nylon, or composite materials thereof. The electrode layer 122 can be made of copper, aluminum, silver, ITO, conductive carbon material, conductive fabric, or flexible conductive film. Further, the surface of the triboelectric layer 121 can be provided with micro / nano protrusions, rough textures, hydrophobic coatings, drainage textures, or replaceable friction films to increase the effective contact area and reduce the impact of water films, mud, or pesticide droplets on the contact separation power generation process. Optionally, the operator can directly touch the triboelectric layer 121 with bare hands, or through thin gloves, conductive gloves, the fingertips of gloves with friction-enhancing pads, or a touch pad driven by the fingers.

[0022] As an exemplary embodiment, the effective touch area of ​​the touch sensing area 11 can be set to 1 cm² to 10 cm², preferably 2 cm² to 6 cm²; the triboelectric layer 121 can be a PTFE, FEP, PDMS, or PVDF film with a thickness of 20 μm to 500 μm; the electrode layer 122 can be made of copper foil, aluminum foil, ITO film, or flexible conductive fabric. The above dimensions and materials are merely examples and do not limit the scope of protection of this invention. Those skilled in the art can select appropriate materials and structures based on touch sensitivity, panel durability, and agricultural field protection requirements.

[0023] like Figure 3As shown, the energy shaping and threshold-triggered delayed light emission branch 13 includes a rectifier circuit 131, an energy storage capacitor 132, a voltage threshold triggering unit 133, a discharge control unit 134, and a low-power LED driving unit 135. The rectifier circuit 131 is used to convert the pulse electrical signal output by the single-electrode triboelectric nanogenerator unit 12 into a unidirectional electrical signal; the energy storage capacitor 132 is used to temporarily store the electrical energy generated by at least one touch action; the voltage threshold triggering unit 133 is used to determine whether the voltage across the energy storage capacitor 132 reaches a preset trigger threshold; the discharge control unit 134 is used to control the energy storage capacitor 132 to release electrical energy to the low-power LED driving unit 135; the low-power LED driving unit 135 is used to drive the corresponding LED light-emitting unit 141 in the LED array 14 to output standardized LED light pulses.

[0024] Because the electrical energy generated by finger contact and separation is characterized by short duration and large amplitude fluctuations, directly driving an LED may result in low-brightness half-pulses, short-term flickering, or missed detection by the camera. Therefore, this invention uses a voltage threshold triggering unit 133 to determine the energy storage state: when the voltage across the energy storage capacitor 132 reaches a preset trigger threshold, the voltage threshold triggering unit 133 is activated, driving the LED light-emitting unit 141 to output standardized LED light pulses; when the energy storage voltage does not reach the preset trigger threshold, the LED light-emitting unit 141 is prohibited from outputting light-encoded signals. This effectively filters out incomplete light pulses caused by weak touches, accidental water droplet contact, leaf contact, mud or dust disturbance, or insufficient energy.

[0025] Furthermore, the duration of the standardized LED light pulse is matched to the frame rate of the visual acquisition module 2. As a preferred embodiment, the duration of the standardized LED light pulse is not less than the time required for the visual acquisition module 2 to continuously acquire N frames of images, where N≥2, preferably N≥3. For example, when the frame rate of the visual acquisition module 2 is 25fps to 60fps, the duration of the standardized LED light pulse can be set to 80ms to 500ms, preferably 100ms to 300ms. The capacity of the energy storage capacitor 132 can be selected from 1μF to 100μF, preferably 10μF to 47μF; the preset trigger threshold can be selected from 2V to 5V. Specific values ​​can be determined based on the output capability of the single-electrode triboelectric nanogenerator 12, the LED load power consumption, the recognition distance, and the frame rate of the visual acquisition module 2. In some embodiments, if the electrical energy generated by a single touch is insufficient to reach the preset trigger threshold, the energy storage capacitor 132 can accumulate the electrical energy generated by multiple touches until the preset trigger threshold is reached before outputting the standardized LED light pulse.

[0026] like Figure 4As shown, the LED array 14 includes multiple LED light-emitting units 141 and at least one verification LED 142. The LED light-emitting units 141 can be arranged in a two-dimensional array, or in a strip, arc, or functional partition manner. The LED array 14 is used to generate a spatial-temporal composite optical coding signal, which includes a device coding segment, an action coding segment, and a verification coding segment. The device coding segment is determined by the spatial position of the LED light-emitting units 141; the action coding segment is determined by the number of flashes, the duration of light emission, the pulse width, or the flash interval; and the verification coding segment is determined by the start pulse, end pulse, repetition pulse, or confirmation pulse of the verification LED 142.

[0027] For example, LED1 represents irrigation equipment, LED2 represents supplemental lighting equipment, LED3 represents ventilation equipment, LED4 represents spraying equipment, LED5 represents fertilization equipment, and LED0 serves as a check LED142. LED1 flashing twice, accompanied by the end of the check pulse from LED0, indicates irrigation is started; LED1 flashing three times, accompanied by the end of the check pulse from LED0, indicates irrigation is stopped; LED2 flashing twice, accompanied by the end of the check pulse from LED0, indicates supplemental lighting is turned on; LED2 flashing three times, accompanied by the end of the check pulse from LED0, indicates supplemental lighting is turned off; LED5 flashing synchronously with LED0 at high frequency indicates an emergency stop. The above encoding method is merely an example, and the present invention does not limit the number of LEDs, their arrangement, or the encoding mapping relationship.

[0028] Furthermore, the LED light-emitting unit 141 in the LED array 14 can be a narrowband LED, and an optical filter matching the emission band of the narrowband LED can be set in front of the visual acquisition module 2. The narrowband LED can be a red LED, blue LED, green LED, or near-infrared LED. By combining the narrowband light source with the filter, interference from sunlight, supplementary lighting, or reflections from the greenhouse film on the recognition results can be reduced. Optionally, the LED array 14 can also be equipped with a light shield, a transparent waterproof window, or a diffuse reflection suppression structure to improve the stability of long-distance visual recognition.

[0029] like Figure 4As shown, the self-generating touch control panel 1 is provided with positioning marks 15. Positioning marks 15 can be high-contrast borders, corner marks, geometric marks, QR code markings, or other positioning structures that can be recognized by the optical encoding recognition module 3. Positioning marks 15 are used to assist the optical encoding recognition module 3 in determining the target recognition area of ​​the LED array 14 in the image. Preferably, positioning marks 15 are located at at least three corner positions around the LED array 14, more preferably at four corner positions. The optical encoding recognition module 3 can calculate the geometric transformation relationship of the LED array area in the image based on the positioning marks 15, such as a perspective transformation matrix, and map the LED array area in the acquired image to a preset standard coordinate system. Therefore, even if there is an installation angle deviation, distance change, or slight panel tilt between the visual acquisition module 2 and the self-generating touch control panel 1, the brightness time series of each LED light-emitting unit 141 can still be stably extracted in the standard coordinate system.

[0030] Furthermore, the optical encoding recognition module 3 can set a point window 7 and a local background window 8 for each LED light-emitting unit 141 in the target recognition area or a preset standard coordinate system. The point window 7 covers the imaging position of the corresponding LED light-emitting unit 141, and the local background window 8 is set around the point window 7 and avoids other LED light-emitting units. The optical encoding recognition module 3 calculates the brightness difference between the average brightness of the point window 7 and the average brightness of the local background window 8, and extracts the brightness time series based on the brightness difference. Compared with directly using an absolute brightness threshold, this local background difference method can adapt to the complex agricultural light environment such as strong daytime light, nighttime supplemental lighting, water mist reflection, greenhouse film reflection, and shadow changes.

[0031] like Figure 5 As shown, the optical encoding recognition module 3 performs multi-frame verification and false trigger suppression on the brightness time series. Specifically, the optical encoding recognition module 3 can determine whether the optical encoding signal is valid based on spatial coordinates, dynamic brightness threshold, number of consecutive frames, flicker interval, encoding segment order, and verification encoding segment. The dynamic brightness threshold can be determined based on the average brightness of the local background window 8, the brightness standard deviation, or the brightness change of historical frames, for example, set as the average brightness of the local background plus a preset multiple of the standard deviation. When the brightness difference between the point window 7 and the local background window 8 exceeds the dynamic brightness threshold and remains valid in images for no less than a preset number of consecutive frames, the optical encoding recognition module 3 determines that the corresponding LED light-emitting unit 141 is in an effective light-emitting state.

[0032] When a candidate light signal fails to meet at least one of the following rules: spatial coordinates, dynamic brightness threshold, consecutive frame count, flicker interval, matching of device code segment and action code segment, or verification code segment, the light code recognition module 3 determines it as an invalid signal and prohibits the instruction parsing module 4 from outputting agricultural equipment control commands. For ordinary control commands, preferably, the control command is only output after the device code segment, action code segment, and verification code segment are all detected to be complete; for emergency stop commands, they can be set to be triggered immediately after a preset high-priority LED combination code is detected in at least two consecutive frames, in order to balance safety and rapid response.

[0033] The instruction parsing module 4 parses the valid spatial-temporal composite optical encoded signal into agricultural equipment control instructions according to a preset mapping relationship. The preset mapping relationship includes the correspondence between LED spatial position and agricultural execution equipment type, the correspondence between flash count or pulse width and control action, the correspondence between verification LED 142 and instruction integrity, and the correspondence between emergency stop combination code and full equipment stop action. The instruction parsing module 4 can be deployed locally in the vision acquisition module 2, on an edge computing terminal, in an agricultural control host, or in an IoT gateway.

[0034] After receiving control commands from agricultural equipment, the agricultural equipment safety control module 5 does not execute them directly. Instead, it makes judgments based on the current operating status of the agricultural execution equipment 6, the environmental status, mutually exclusive command relationships, control command priorities, and preset safety conditions. For example, when the command parsing module 4 outputs an irrigation start command, the agricultural equipment safety control module 5 determines whether the water pump is functioning normally, whether the solenoid valve is available, whether the water source level meets requirements, and whether the soil moisture exceeds a preset upper limit. If the safety conditions are met, it controls the water pump to start and opens the solenoid valve; if the safety conditions are not met, it refuses to execute the irrigation start command. Similarly, when an additional lighting start command is output, the agricultural equipment safety control module 5 can estimate the brightness using a light sensor or visual estimation to determine whether the current ambient light is below a preset threshold, and can prohibit additional lighting start if the ambient temperature exceeds a preset high-temperature threshold. Furthermore, when a spraying start command is output, the agricultural equipment safety control module 5 can determine whether there are personnel in the spraying area, whether the ventilation equipment is in a mutually exclusive state, or whether the spraying equipment is within a cooling interval; if a safety risk exists, it refuses to execute the spraying start command. For roller shutter equipment or greenhouse window opening mechanisms, the agricultural equipment safety control module 5 can determine whether to allow the corresponding action based on the wind and rain conditions, window position, or roller shutter operation status.

[0035] In some implementations, the agricultural equipment safety control module 5 is also connected to an equipment status feedback module to acquire the start / stop status, current status, fault status, valve opening / closing status, or execution feedback status of the agricultural execution equipment 6. When the agricultural equipment safety control module 5 issues a control command, if the agricultural execution equipment 6 fails to respond within a preset time, the system can output an abnormality prompt, record fault information, cancel the control command, or execute a protection action. Preferably, the emergency stop command has a higher priority than other control commands. When the optical code recognition module 3 recognizes the high-priority combination code corresponding to the emergency stop, the agricultural equipment safety control module 5 can control the water pump, solenoid valve, fan, supplemental lighting, spraying equipment, fertilizer pump, and nutrient solution circulation pump to enter a safe stop state.

[0036] This invention utilizes the contact and separation action between a finger and the triboelectric layer 121 to achieve self-generation of the touch terminal. Through threshold triggering and frame rate matching delayed light emission, random transient electrical energy is converted into standardized LED light pulses. A spatial-temporal composite optical coding protocol is constructed through device coding segment, action coding segment, and verification coding segment. Stable recognition in complex agricultural light environments is achieved through positioning marker 15, target recognition area, brightness time series, and multi-frame verification. The risk of erroneous execution is reduced through agricultural equipment safety control module 5. It is suitable for greenhouses, plant factories, smart farmland, orchards, breeding farms, and agricultural robot operating environments.

Claims

1. A smart agricultural vision control system based on finger-touch self-generated standardized optical pulse coding, characterized in that, include: The self-generating touch control panel includes multiple touch sensing areas, multiple single-electrode triboelectric nanogenerators, energy shaping and threshold-triggered delay light-emitting branches, LED arrays and positioning markers. Each touch sensing area corresponds to at least one single-electrode triboelectric nanogenerator unit. The single-electrode triboelectric nanogenerator unit includes a triboelectric layer disposed on the surface of the touch sensing area and an electrode layer disposed below the triboelectric layer. When the operator's finger or touch contact body touches and leaves the triboelectric layer, the electrode layer outputs transient pulse electrical energy. The energy shaping and threshold-triggered delayed light emission branch is electrically connected to the electrode layer and the LED array. It is used to rectify and store the transient pulse energy, and drive at least one LED light-emitting unit in the LED array to output a standardized LED light pulse when the energy storage voltage reaches a preset trigger threshold. The duration of the standardized LED light pulse is matched with the acquisition frame rate of the visual acquisition module so that the standardized LED light pulse maintains detectable brightness in multiple consecutive frames of images acquired by the visual acquisition module. The LED array is used to generate a spatial-temporal composite optical coding signal, which includes a device coding segment, an action coding segment, and a verification coding segment. The device coding segment is determined by the spatial position of the LED light-emitting unit. The action coding segment is determined by the number of flashes, the duration of light emission, the pulse width, or the flash interval of the LED light-emitting unit. The verification coding segment is determined by the start pulse, end pulse, or acknowledgment pulse of at least one verification LED. The visual acquisition module is used to acquire images or video sequences containing the LED array; The optical coding recognition module is used to determine the target recognition area of ​​the LED array according to the positioning mark, extract the brightness time series of each LED light-emitting unit within the target recognition area, and determine whether the spatial-temporal composite optical coding signal is valid according to the spatial position, dynamic brightness threshold, number of consecutive frames, time series rules and verification coding segment. The instruction parsing module is used to parse valid spatial-temporal composite optical coded signals into agricultural equipment control instructions; The agricultural equipment safety control module is used to control the operation of the agricultural equipment or prohibit the execution of corresponding agricultural equipment control commands based on the agricultural equipment control commands, the current operating status of the agricultural execution equipment, the mutual exclusion command relationship, the control command priority, or preset safety conditions.

2. The intelligent agricultural vision control system based on finger-touch self-generated standardized optical pulse coding according to claim 1, characterized in that, The energy shaping and threshold-triggered delayed light emission branch includes a rectifier circuit, an energy storage capacitor, a voltage threshold triggering unit, a voltage limiting unit, a discharge control unit, and a low-power LED driving unit. When the voltage across the energy storage capacitor reaches a preset trigger threshold, the voltage threshold triggering unit turns on the low-power LED driving unit. When the energy storage voltage does not reach the preset trigger threshold, the LED light emission unit is prohibited from outputting light encoding signals. The duration of the standardized LED light pulse is not less than the time required for the visual acquisition module to continuously acquire N frames of images, where N≥2.

3. The intelligent agricultural vision control system based on finger-touch self-generated standardized light pulse coding according to claim 1, characterized in that, The positioning markers include at least three corner markers, a high-contrast border, geometric markers, or QR code calibration markers set around the LED array; the optical coding recognition module determines the geometric transformation relationship of the LED array area based on the positioning markers and maps the LED array area in the acquired image to a preset standard coordinate system; the optical coding recognition module sets a point window and a local background window for each LED light-emitting unit in the target recognition area or the preset standard coordinate system, and extracts the brightness time series based on the difference between the brightness of the point window and the brightness of the local background window.

4. The intelligent agricultural vision control system based on finger-touch self-generated standardized light pulse coding according to claim 1, characterized in that, The surface of the triboelectric layer is provided with micro-nano protrusions, rough textures, hydrophobic coatings, drainage textures, or replaceable friction films; the touch contact includes bare fingers, fingers wearing conductive gloves, glove fingertips with friction-enhancing sheets, or touch sheets driven by fingers; the LED light-emitting unit is a narrowband LED, and an optical filter matching the emission band of the narrowband LED is provided in front of the visual acquisition module.

5. The intelligent agricultural vision control system based on finger-touch self-generated standardized optical pulse coding according to claim 1, characterized in that, The agricultural execution equipment includes at least one of the following: water pump, solenoid valve, fan, supplemental lighting, spraying equipment, fertilizer pump, nutrient solution circulation pump, roller shutter equipment, wet curtain, greenhouse window opening mechanism, or agricultural robot; the agricultural equipment safety control module sets the emergency stop command to have a higher priority than other agricultural equipment control commands; before executing the irrigation start command, the agricultural equipment safety control module determines whether the water pump status, solenoid valve status, water source level, or soil moisture meet preset safety conditions; before executing the supplemental lighting start command, it determines whether the ambient light or ambient temperature meets preset safety conditions; and before executing the spraying start command, it determines whether the personnel status, ventilation equipment status, or spraying cooling interval in the spraying area meet preset safety conditions.

Images