Intelligent real-time image defogging helmet for harsh dust fog environment and method of use thereof

Abstract

The application provides an intelligent real-time image defogging helmet for a severe dust and fog environment and a use method thereof, and can realize real-time dust and fog removal processing of images in a severe dust and fog environment. Through processing of images in collected video information, real-time dust and fog removal processing of images in a severe dust and fog environment can be realized, work and life of people in the severe dust and fog environment are facilitated, visual visibility is improved, survival probability is ensured, and a stable and safe work and survival environment is maintained. Meanwhile, low power consumption and green environmental protection are taken as cores, human-machine collaborative development is promoted, the helmet is intelligentized, and the application prospect is wide.

Description

Technical Field

[0001] This invention relates to the field of helmet and image processing technology, specifically to an intelligent real-time image defogging helmet for harsh dust and fog environments and its usage method. Background Technology

[0002] In their daily work and travel, people are often troubled by harsh dust and fog environments, such as quarry workers and ordinary people in foggy areas. Quarry environments contain a large amount of dust, which can easily cause injury to workers when performing high-precision operations such as cutting stones. In addition, in recent years, many areas have experienced persistent smog, leading to frequent traffic accidents, which are particularly dangerous for non-motorized vehicle owners and pedestrians.

[0003] Existing helmets primarily focus on protecting the brain, but they are ineffective in such harsh environments and cannot address the issue of visual clarity to ensure people's safety. Other smart helmets are equipped with cameras, microphones, and wireless communication modules, but they lack the ability to process images in real time using the helmet's built-in computing devices. Summary of the Invention

[0004] In view of this, the present invention proposes an intelligent real-time image defogging helmet for harsh dust and fog environments and its usage method, which can realize real-time dust and fog defogging processing of images in harsh dust and fog environments.

[0005] To achieve the above objectives, the technical solution of the present invention is as follows:

[0006] This invention discloses an intelligent real-time image defogging helmet for harsh dust and fog environments, comprising an intelligent helmet body and an intelligent device client. The intelligent helmet body includes a data acquisition module, a processing module, a communication module, a display module, and a power supply module. The intelligent device client is used to manage the intelligent helmet, pair the intelligent helmet, record the location information of the intelligent helmet, adjust the parameters of the displayed image, and set the user's emergency contact.

[0007] The acquisition module includes a visible light camera, a GPS positioning module, an inertial measurement unit, and a touchpad. The visible light camera is fixed to the front of the helmet shell to acquire video information in the current environment. The GPS positioning module and the inertial measurement unit are fixed in the space between the helmet shell and the inner layer of the helmet to acquire the helmet's current position, attitude angle, and acceleration, respectively. The touchpad is fixed to the left edge of the helmet shell to acquire the user's touch information.

[0008] The processing module is fixed between the helmet shell and the inner layer of the helmet. It is used to acquire video information in the current environment, restore dust-free fog images through a defogging algorithm based on an integrated defogging network, receive parameter information from the smart device client, and adjust the parameters of the image.

[0009] The communication module is fixed between the helmet shell and the helmet inner layer, and is connected to the processing module. It is used to send the information collected by the acquisition module to the smart device client. The smart device client sends the added or modified information back to the processing module.

[0010] The display module is fixed to the helmet body and is used to display a dust-free, fog-free image after the processing module has processed the real image of the current environment.

[0011] The power supply module is fixed to the helmet shell and provides power to the acquisition module, processing module and display module.

[0012] The defogging algorithm of the integrated defogging network is an improved AODNet defogging algorithm, which includes five convolutional layers, three connection layers, histogram equalization grayscale transformation method and Laplacian operator filter.

[0013] The smart device client includes a Bluetooth serial communication module for receiving data from the processing module and sending data back to the processing module; it also has a storage function to record the helmet's real-time position; and a notification function that sends an SMS to emergency contacts when the data collected by the inertial measurement unit changes abruptly.

[0014] The processing module sends a danger signal to the smart device client connected to the communication module; after receiving the danger signal, the smart device client sends a text message with GPS location information to the user's emergency contact.

[0015] The present invention discloses a method for using an intelligent real-time image defogging helmet for harsh dust and fog environments, comprising the following specific steps:

[0016] Step 1: Training is based on the improved AODNet dehazing algorithm. The training set of the network consists of a public dataset and 200 hazy and sunny day images collected by volunteers at the same location using the same shooting method. The acquired images are processed through five convolutional layers, three connection layers, histogram equalization grayscale transformation method and Laplacian operator filter to obtain dehazed images.

[0017] Step 2: Port the trained network model algorithm into the software of the smart helmet, distribute the smart helmets to people in harsh dust and fog environments, and have users wear the smart helmets. If they have smart devices, install the smart device client described in this invention. Users can enter their personal emergency contact information on the smart device client.

[0018] After the user turns on the power switch on the smart helmet, the smart helmet uses a visible light camera to capture the environment in front of the user's field of vision. The processing module uses an improved AODNet dehazing algorithm to perform real-time image dehazing. After image dehazing, the image is sent to the display module and displayed in front of the user.

[0019] Step 3: The processing module receives GPS positioning information, attitude information and acceleration collected by the inertial measurement unit from the smart helmet worn by the user in real time; by analyzing the attitude information and acceleration, it determines the user's current status and safety situation; if the data changes abruptly or exceeds the threshold, it is analyzed that the user is in danger, and the processing module sends a signal to the smart device client via Bluetooth, and the smart device client sends a call for help to emergency contacts with GPS positioning information.

[0020] Step 4: Depending on the severe dust and fog conditions within the field of vision, the user can generate a sliding signal by sliding their finger on the resistive touchscreen of the smart helmet, or adjust the image quality by inputting parameters through the smart device client. Depending on the dangerous situation, the user can send a distress signal by tapping the capacitive touchscreen. The resistive touchscreen sends different digital signals to the processing module based on the different touch methods. When the processing module receives the sliding signal or input parameters from the smart device client, it adjusts the parameters in the Laplacian operator filter based on the improved AODNet defogging algorithm. When a distress signal is received, it is sent to the smart device client via Bluetooth, and the smart device client sends a distress signal with GPS location information to emergency contacts.

[0021] Beneficial effects:

[0022] The helmet of this invention processes images from captured video information, enabling real-time de-dust and de-fogging of images in harsh dusty and foggy environments. This provides convenience for people working and living in such environments, improves visual visibility, increases the chances of survival, and maintains a stable and safe working and living environment. Furthermore, with low power consumption and environmental friendliness as its core features, it promotes human-machine collaboration, realizes helmet intelligence, and has broad application prospects.

[0023] The drawback of the original AODNet dehazing algorithm was the presence of distortion and blurring after image processing. This invention employs an improved AODNet dehazing algorithm that adds two modules: histogram equalization grayscale transformation and a Laplacian operator filter. Histogram equalization grayscale transformation enhances the overall image contrast by altering the grayscale levels of individual pixels. The Laplacian operator filter enhances areas of abrupt grayscale changes and weakens areas of slow grayscale changes, aiming to extract edge information and sharpen blurred images. The improved AODNet dehazing algorithm retains the lightweight advantage of the original algorithm while achieving realistic and clear color reproduction.

[0024] This invention, through image processing of acquired video information, enables real-time de-dust and de-fogging of images in harsh dusty and foggy environments. This provides convenience for people working and living in such environments, improves visual visibility, enhances survival chances, and maintains a stable and safe working and living environment. Furthermore, with low power consumption and environmental friendliness as its core principles, it promotes human-machine collaboration, realizes helmet intelligence, and has broad application prospects. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the external structure of the intelligent real-time image defogging helmet for harsh dust and fog environments proposed in this invention.

[0026] Figure 2 This is a schematic diagram of the internal structure of the intelligent real-time image defogging helmet for harsh dust and fog environments proposed in this invention.

[0027] Figure 3 This is a system connection flowchart of the present invention.

[0028] Figure 4 This is a flowchart of the algorithm for the intelligent real-time image defogging helmet for harsh dust and fog environments proposed in this invention.

[0029] Among them, 1-visible light camera; 2-miniature display screen; 3-touchpad; 4-solar cell; 5-overall internal module. Detailed Implementation

[0030] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0031] This invention relates to an intelligent real-time image defogging helmet for harsh dust and fog environments. It enables real-time defogging of images in such conditions, offering fast processing speed and clear image quality. The algorithm used in this invention is compatible with various devices. The helmet comprises a smart helmet body and a smart device client. The smart helmet body includes a data acquisition module, a processing module, a communication module, a display module, and a power supply module; it is used to manage the smart helmet, pair with other smart helmets, record the helmet's location information, adjust display image parameters, and set the user's emergency contacts. Figure 1 This is a schematic diagram of the external structure of the intelligent real-time image defogging helmet for harsh dust and fog environments proposed in this invention. Figure 2 This is a schematic diagram of the internal structure of the intelligent real-time image defogging helmet for harsh dust and fog environments proposed in this invention. Figure 3 This is a system connection flowchart of the present invention.

[0032] Specifically, the acquisition module includes a visible light camera, a GPS positioning module, an inertial measurement unit, and a touchpad. The visible light camera is fixed to the front of the helmet shell to acquire video information in the current environment. The GPS positioning module and the inertial measurement unit are fixed in the space between the helmet shell and the inner layer of the helmet to acquire the helmet's current position, attitude angle, and acceleration, respectively. The touchpad is fixed to the left edge of the helmet shell to acquire the user's touch information, with different touch methods representing different command information.

[0033] The processing module is fixed between the helmet shell and the inner layer of the helmet. It is used to acquire video information in the current environment, restore the dust-free fog image through the image defogging processing stage based on the improved integrated defogging network (AODNet), receive parameter information from the smart device client, and adjust the image parameters. In this embodiment, the processing module adopts a small computing device.

[0034] The communication module is fixed between the helmet shell and the inner layer of the helmet and is connected to the processing module. It is used to send the information collected by the acquisition module (such as GPS positioning, reminder instructions, etc.) to the smart device client. The smart device client sends the added or modified information back to the processing module. In this embodiment, Bluetooth is used for communication.

[0035] The display module is fixed to the helmet body and is used to display a dust-free, fog-free image of the current environment after processing by the processing module; in this embodiment, the display module is a miniature display screen.

[0036] The power supply module is fixed to the helmet shell and provides energy to the data acquisition module, processing module, and display module. In this embodiment, the power supply module uses polymer batteries and solar cells, and adopts an integrated structure of polymer batteries and solar rechargeable batteries to achieve self-charging and discharging functionality.

[0037] All modules in the acquisition module can transmit data to the processing module via a data cable. The processing module then transmits the relevant information to the miniature display screen via the data cable to enable real-time playback of the color video.

[0038] Furthermore, the smart device client includes a Bluetooth serial communication module for receiving data information (such as GPS positioning, reminder commands, etc.) sent by the processing module and sending data to the processing module; it also includes a storage function to record the helmet's real-time location; and a reminder function, which sends an SMS to emergency contacts when the data collected by the inertial measurement unit changes abruptly.

[0039] The processing module sends a danger signal to the smart device client connected via the communication module. Upon receiving the danger signal, the smart device client sends a text message containing GPS location information to the user's emergency contact.

[0040] Furthermore, the image dehazing process based on the improved AODNet dehazing algorithm in the processing module includes five convolutional layers, three connection layers, histogram equalization grayscale transformation, and a Laplacian operator filter. The image after dehazing is displayed on a miniature display screen.

[0041] This invention also provides a method for using an intelligent real-time image defogging helmet for harsh dust and fog environments. Using the helmet of this invention includes the following specific steps:

[0042] Step 1: Training based on the improved AODNet dehazing algorithm. The network training set consists of a public dataset and 200 images of hazy and sunny weather collected by volunteers at the same location using the same shooting methods. The network flow is as follows: Figure 4 As shown, the acquired image is processed through five convolutional layers, three connection layers, histogram equalization grayscale transformation, and Laplacian operator filter to obtain a dehazed image.

[0043] Step 2: Port the trained network model algorithm into the smart helmet's software, and it can then be officially put into use and implement related functions. Distribute the smart helmets to people in harsh dusty and foggy environments, and instruct users to wear them correctly. If they have a smart device, they can install the smart device client described in this invention. After the user turns on the power switch on the smart helmet, the helmet uses a visible light camera to capture images of the environment in front of the user's field of vision. The processing module uses an improved AODNet dehazing algorithm to perform real-time image dehazing. The dehazed image is then transmitted to a miniature display screen and displayed in front of the user. Users also need to enter their personal emergency contact information on the smart device client for emergency assistance.

[0044] Step 3: The processing module receives real-time GPS positioning information from the user's smart helmet, attitude information collected by the inertial measurement unit, and acceleration data. By analyzing the attitude and acceleration information, it determines the user's current state and safety. If the data shows a sudden change or exceeds a threshold, it is analyzed that the user is in danger. The processing module sends a signal to the smart device client via Bluetooth, and the smart device client sends a distress signal containing GPS positioning information to emergency contacts.

[0045] Step 4: Depending on the severe dust and fog conditions within the field of vision, the user can generate a sliding signal by sliding their finger on the resistive touchscreen of the smart helmet, or adjust the image quality through the input parameter function in the smart device client. Depending on the dangerous situation in the current environment, the user can send a distress signal by tapping the capacitive touchscreen. The resistive touchscreen sends different digital signals to the processing module based on different touch methods. When the processing module receives a sliding signal or input parameters from the smart device client, it adjusts the parameters in the Laplacian operator filter based on the improved AODNet defogging algorithm to improve image quality; when it receives a distress signal, it sends it to the smart device client via Bluetooth, and the smart device client sends a distress signal with GPS location information to emergency contacts.

[0046] In summary, the above are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A smart real-time image defogging helmet for harsh dust and fog environments, characterized in that, The system includes a smart helmet body and a smart device client. The smart helmet body includes a data acquisition module, a processing module, a communication module, a display module, and a power supply module. The smart device client is used to manage the smart helmet, pair the smart helmet, record the location information of the smart helmet, adjust the parameters of the displayed image, and set the user's emergency contacts. The acquisition module includes a visible light camera, a GPS positioning module, an inertial measurement unit, and a touchpad. The visible light camera is fixed to the front of the helmet shell to acquire video information in the current environment. The GPS positioning module and the inertial measurement unit are fixed in the space between the helmet shell and the inner layer of the helmet to acquire the helmet's current position, attitude angle, and acceleration, respectively. The touchpad is fixed to the left edge of the helmet shell to acquire the user's touch information. The processing module is fixed between the helmet shell and the inner layer of the helmet. It is used to acquire video information in the current environment, restore dust-free fog images through a defogging algorithm based on an integrated defogging network, receive parameter information from the smart device client, and adjust the parameters of the image. The communication module is fixed between the helmet shell and the helmet inner layer, and is connected to the processing module. It is used to send the data information of the processing module to the smart device client. The smart device client will send the added or modified information back to the processing module. The display module is fixed to the helmet body and is used to display a dust-free, fog-free image after the processing module has processed the real image of the current environment. The power supply module is fixed to the helmet shell and provides power to the data acquisition module, processing module and display module; The defogging algorithm of the integrated defogging network is the improved AODNet defogging algorithm, which includes five convolutional layers, three connection layers, histogram equalization grayscale transformation method and Laplacian operator filter; The process of using the intelligent real-time image defogging helmet includes: Step 1: Training is based on the improved AODNet dehazing algorithm. The training set of the network consists of a public dataset and 200 hazy and sunny day images collected by volunteers at the same location using the same shooting method. The acquired images are processed sequentially through five convolutional layers, three connection layers, histogram equalization grayscale transformation method and Laplacian operator filter to obtain dehazed images. Step 2: Port the trained network model algorithm into the software of the smart helmet, distribute the smart helmets to people in harsh dust and fog environments, and have users wear the smart helmets. If they have smart devices, install the smart device client described in this invention. Users can enter their personal emergency contact information on the smart device client. After the user turns on the power switch on the smart helmet, the smart helmet uses a visible light camera to capture the environment in front of the user's field of vision. The processing module uses an improved AODNet dehazing algorithm to perform real-time image dehazing. After image dehazing, the image is sent to the display module and displayed in front of the user. Step 3: The processing module receives GPS positioning information, attitude information and acceleration collected by the inertial measurement unit from the smart helmet worn by the user in real time; by analyzing the attitude information and acceleration, it determines the user's current status and safety situation; if the data changes abruptly or exceeds the threshold, it is analyzed that the user is in danger, and the processing module sends a signal to the smart device client via Bluetooth, and the smart device client sends a call for help to emergency contacts with GPS positioning information. Step 4: Based on the severe dust and fog conditions within the field of vision, the user generates a sliding signal by sliding their finger on the resistive touchscreen of the smart helmet; depending on the current dangerous situation, the user sends a distress signal by tapping the capacitive touchscreen; the resistive touchscreen sends different digital signals to the processing module according to different touch methods; when the processing module receives the sliding signal, it adjusts the parameters in the Laplace operator filter based on the improved AODNet defogging algorithm; when a distress signal is received, it is sent to the smart device client via Bluetooth module, and the smart device client sends a distress signal with GPS location information to emergency contacts. The smart device client includes a notification function. When the data collected by the inertial measurement unit changes abruptly, the smart device client will send an SMS to the emergency contact.

2. The helmet as described in claim 1, characterized in that, The smart device client includes a Bluetooth serial communication module for receiving data from the processing module and sending data to the processing module; it also includes a storage function to record the helmet's real-time position.

Images