KR20220102294A - Transmitter recognition apparatus and method for optical camera communication between vehicle and infrastructure - Google Patents

Abstract

The present invention relates to a device and method for recognizing a transmitter of an optical camera communication between a vehicle and an infrastructure. A method for recognizing a transmitter includes the steps of: receiving image data for a transmitter on a road infrastructure; configuring a data set by matching a bounding box for the transmitter on the road infrastructure and classification information for the transmitter, setting a bounding box only for a light source in the transmitter; learning an YOLO model using the data set; and recognizing a bounding box output by processing image data for road infrastructure input to the learned YOLO model as the transmitter of optical camera communication. Therefore, in vehicle-infrastructure optical camera communication, it is possible to accurately detect the transmitter on the infrastructure in real time.

Description

Transmitter recognition device and method for vehicle-infrastructure optical camera communication

The present specification relates to an apparatus and method for recognizing a transmitter in vehicle-infrastructure (V2I) optical camera communication (OCC).

Although recently sold cars use sensors such as cameras, lidar, and radar to recognize the surrounding environment and assist the driver in driving safely, there are situations in which it is difficult to check the surrounding environment of the vehicle only with the sensors installed in the vehicle. have. Even in such a situation, information about the traffic situation is exchanged through wireless communication (V2I communication) between the vehicle and road infrastructure in order to assist the safe driving of the vehicle, and through this, information about the vehicle's surrounding environment, which is difficult to determine only with the vehicle's sensor, is collected. technology is being researched. In particular, vehicle-infrastructure (V2I) communication is being studied as a key element technology of autonomous driving technology of automobiles, and optical camera communication (OCC) technology is used as a vehicle-infrastructure (V2I) communication means. method is being studied.

Korean Patent Publication No. 10-1971370, 2019.04.16.

In optical camera communication, LED lighting is used as a transmitter, a camera is used as a receiver, and the visible light emitted by LED lighting is used as a communication medium. is used, and as the transmitter LED lighting, traffic lights, street lights, traffic information billboards, and advertisement boards among various road infrastructures may be used.

However, if the transmitter is detected by using the differential image of the transmitter captured by the vehicle, the data reception rate is reduced because the setting and tracking of the region of interest for data extraction becomes unstable due to a sudden change in the position of the LED transmitter according to the movement of the vehicle. There is a problem of reduction.

Therefore, for efficient optical camera communication in vehicle-infrastructure (V2I) communication, there is an increasing need for a technology for accurately tracking and extracting LED lights on road infrastructures, which are transmitters, at high speed.

The problems to be solved by the present invention are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the following description.

The present specification proposes a transmission unit recognition method performed by a transmission unit recognition device for vehicle-to-infrastructure optical camera communication. The transmitter recognition method includes receiving image data for a transmitter on a road infrastructure, matching a bounding box for a transmitter on the road infrastructure and classification information for the transmitter to configure a data set, but in the transmitter Setting the bounding box only for the light source part, learning the YOLO model using the data set, and processing the image data for the road infrastructure input to the learned YOLO model and outputting the bounding box to the optical camera communication It may include the step of recognizing as a transmitter of

The transmitter recognition method and other embodiments may include the following features.

According to an embodiment, a data set is configured by matching a bounding box for a transmitter on the road infrastructure and classification information for the transmitter, but setting a bounding box only for a light source portion in the transmitter includes: It may include labeling the transmitter in the image data to update a weight for detecting the transmitter.

In addition, according to an embodiment, a data set is configured by matching a bounding box for a transmitter on the road infrastructure and classification information for the transmitter, but setting a bounding box for only the light source part in the transmitter includes: , setting an area including only the LED light source part in the entire image for the transmitter as a bounding box.

Also, according to an embodiment, the method for recognizing the transmitter may further include extracting an optical camera communication signal from a light source in the bounding box.

On the other hand, the present specification presents an apparatus for recognizing a transmission unit of an optical camera communication between a vehicle and an infrastructure. The transmitter recognition apparatus configures a data set by matching a data receiver for receiving image data for a transmitter on the road infrastructure, a bounding box for the transmitter on the road infrastructure, and classification information for the transmitter, the transmitter A training data generation unit that sets a bounding box only for the light source in the It may include a transmitter recognition unit for recognizing the bounding box to be as a transmitter of optical camera communication.

The transmitter recognition apparatus and other embodiments may include the following features.

According to an embodiment, the training data generator may update the weight for detecting the transmitter by labeling the transmitter in the image data.

In addition, according to an embodiment, the learning data generator may set a region including only the LED light source part in the entire image of the transmitter as a bounding box.

Also, according to an embodiment, the transmitter recognition unit may extract an optical camera communication signal from a light source in the bounding box.

On the other hand, the present specification proposes a transmission unit recognition method performed by a transmission unit recognition device of vehicle-infrastructure optical camera communication. The transmitter recognition method comprises the steps of receiving image data for a transmitter on the road infrastructure and constructing a data set for learning the YOLO model, training the YOLO model using the data set, and inputting the learned YOLO model Outputting the recognition result for the transmitter on the road infrastructure by processing the image data, but may include setting and outputting a bounding box only for the light source in the entire image of the recognized transmitter.

According to the embodiment disclosed in this specification, there is an effect of accurately detecting a transmitter on an infrastructure in real time in vehicle-infrastructure-to-infrastructure optical camera communication.

In addition, according to the embodiment disclosed in the present specification, in the optical camera communication between vehicle-infrastructure, only the portion where light is actually output from the transmission unit on the infrastructure can be accurately detected in real time, thereby increasing the data reception rate.

On the other hand, the effects obtainable in the present invention are not limited to the above-mentioned effects, and other effects not mentioned will be clearly understood by those of ordinary skill in the art to which the present invention belongs from the description below. will be able

The following drawings attached to the present specification illustrate preferred embodiments of the present invention, and serve to further understand the technical spirit of the present invention together with specific details for carrying out the invention, so that the present invention is described in such drawings. It should not be construed as being limited only to the matters.
1 shows a perceptron structure to which the method proposed in the present specification can be applied.
2 shows a multi-layered perceptron structure to which the method proposed in the present specification can be applied.
3 shows an example of a deep neural network to which the method proposed in the present specification can be applied.
4 shows an example of a convolutional neural network to which the method proposed in this specification can be applied.
5 illustrates a filter operation in a convolutional neural network to which the method proposed in this specification can be applied.
6 illustrates a neural network structure in which a cyclic loop to which the method proposed in the present specification can be applied.
7 shows an operation structure of a recurrent neural network to which the method proposed in the present specification can be applied.
8 is an example of a DNN model to which the method proposed in the present specification can be applied.
9 is a block diagram of an AI device according to an embodiment of the present invention.
10 shows the concept of vehicle-to-infrastructure optical camera communication.
11 illustrates the concept of a vehicle-infrastructure optical camera communication system based on YOLO detection according to an embodiment of the present invention.
12 is a diagram illustrating an apparatus for recognizing a transmitter in vehicle-infrastructure-to-infrastructure optical camera communication according to an embodiment of the present invention.
13 shows an example of transmitter recognition, and shows an example of a bounding box output result for a transmitter by the YOLO model.
14 shows an example of the recognition result for the transmitter by the YOLO model in the apparatus for recognizing the transmitter according to an embodiment of the present invention. An example is shown, and FIG. 14(b) shows an example of a bounding box output result for the transmitter when the transmitter is a street lamp.
15 illustrates a method for detecting a transmitter of an apparatus for recognizing a transmitter according to an embodiment of the present invention.

The technology disclosed herein may be applied to a vehicle-to-infrastructure communication technology based on optical camera communication. However, the technology disclosed in the present specification is not limited thereto, and may be applied to all devices and methods to which the technical idea of the technology may be applied.

It should be noted that the technical terms used in this specification are only used to describe specific embodiments, and are not intended to limit the spirit of the technology disclosed herein. In addition, the technical terms used in this specification should be interpreted in the meaning generally understood by those of ordinary skill in the art to which the technology disclosed in this specification belongs, unless otherwise defined in this specification. It should not be construed in an overly comprehensive sense or in an excessively narrow sense. In addition, when the technical term used in this specification is an erroneous technical term that does not accurately express the spirit of the technology disclosed in this specification, a technical term that can be correctly understood by those of ordinary skill in the art to which the technology disclosed in this specification belongs. should be understood and replaced with In addition, general terms used in this specification should be interpreted according to the definition in the dictionary or contextual context, and should not be interpreted in an excessively reduced meaning.

As used herein, terms including ordinal numbers such as first, second, etc. may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, a first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component.

Hereinafter, the embodiments disclosed in the present specification will be described in detail with reference to the accompanying drawings, but the same or similar components are assigned the same reference numerals regardless of reference numerals, and redundant description thereof will be omitted.

In addition, in describing the technology disclosed in this specification, if it is determined that a detailed description of a related known technology may obscure the gist of the technology disclosed in this specification, the detailed description thereof will be omitted. In addition, it should be noted that the accompanying drawings are only for easy understanding of the spirit of the technology disclosed in this specification, and should not be construed as limiting the spirit of the technology by the accompanying drawings.

Hereinafter, artificial intelligence learning that can be applied to the optical camera communication between the vehicle-infrastructure of the present invention will be described with reference to the accompanying drawings 1 to 8 .

Artificial intelligence (hereinafter referred to as AI) can use numerous analytics to determine how complex target tasks are performed. In other words, AI can increase efficiency and reduce processing delays. Time-consuming tasks such as analyzing large amounts of data can be performed instantly by using AI.

Hereinafter, machine learning, a kind of AI, will be looked at in more detail.

Machine learning refers to a set of actions that trains a machine to create a machine that can perform tasks that humans can or cannot do. Machine learning requires data and a learning model. In machine learning, data learning methods can be roughly divided into three types: supervised learning, unsupervised learning, and reinforcement learning.

Neural network learning is to minimize errors in output. Neural network learning repeatedly inputs learning data to the neural network, calculates the output and target errors of the neural network for the training data, and backpropagates the neural network error from the output layer of the neural network to the input layer in the direction to reduce the error. ) to update the weights of each node in the neural network.

Supervised learning uses training data in which the correct answer is labeled in the training data, and in unsupervised learning, the correct answer may not be labeled in the training data. That is, for example, the training data in the case of supervised learning related to data classification may be data in which categories are labeled for each of the training data. The labeled training data is input to the neural network, and an error can be calculated by comparing the output (category) of the neural network with the label of the training data. The calculated error is back propagated in the reverse direction in the neural network (ie, from the output layer to the input layer), and the connection weight of each node of each layer of the neural network may be updated according to the back propagation. A change amount of the connection weight of each node to be updated may be determined according to a learning rate. The computation of the neural network on the input data and the backpropagation of errors can constitute a learning cycle (epoch). The learning rate may be applied differently depending on the number of repetitions of the learning cycle of the neural network. For example, in the early stages of learning of a neural network, a high learning rate can be used to increase the efficiency by allowing the neural network to quickly obtain a certain level of performance, and in the late learning period, a low learning rate can be used to increase the accuracy.

The learning method may vary depending on the characteristics of the data. For example, in a case where the purpose of accurately predicting data transmitted from a transmitter in a communication system is accurately predicted by a receiver, it is preferable to perform learning using supervised learning rather than unsupervised learning or reinforcement learning.

The learning model corresponds to the human brain, and the most basic linear model can be considered, but the paradigm of machine learning that uses high-complexity neural network structures such as artificial neural networks (ANN) as the learning model is called deep learning. called deep learning).

The neural network cord used as a learning method is largely divided into deep neural networks (DNN), convolutional deep neural networks (CNN), and recurrent neural networks (RNN). have.

An artificial neural network is an example of connecting several perceptrons.

1 shows a perceptron structure to which the method proposed in the present specification can be applied.

Referring to FIG. 1 , when an input vector x=(x ₁ , x ₂ , ... , x _d ) is input, each component is multiplied by a weight (W ₁ , W ₂ , ... , W _d ), and the result After summing all the , the whole process of applying the activation function σ(·) is called a perceptron. The huge artificial neural network structure may extend the simplified perceptron structure shown in FIG. 1 to apply input vectors to different multidimensional perceptrons. For convenience of description, an input value or an output value is referred to as a node.

Meanwhile, the perceptron structure shown in FIG. 1 can be described as being composed of a total of three layers based on an input value and an output value. An artificial neural network in which H (d+1)-dimensional perceptrons exist between the 1st layer and the 2nd layer and K (H+1)-dimensional perceptrons exist between the 2nd layer and the 3rd layer can be expressed as shown in FIG. 2 .

2 shows the structure of a multi-layer perceptron to which the method proposed in the present specification can be applied.

The layer where the input vector is located is called the input layer, the layer where the final output value is located is called the output layer, and all the layers located between the input layer and the output layer are called the hidden layer. In the example of FIG. 2 , three layers are disclosed, but when counting the actual number of artificial neural network layers, the input layer is counted excluding the input layer, so it can be viewed as a total of two layers. The artificial neural network is constructed by connecting the perceptrons of the basic blocks in two dimensions.

The aforementioned input layer, hidden layer, and output layer can be jointly applied in various artificial neural network structures such as CNN and RNN to be described later as well as multi-layer perceptron. As the number of hidden layers increases, the artificial neural network becomes deeper, and a machine learning paradigm that uses a sufficiently deep artificial neural network as a learning model is called deep learning. Also, an artificial neural network used for deep learning is called a deep neural network (DNN).

3 shows an example of a deep neural network to which the method proposed in the present specification can be applied.

The deep neural network shown in FIG. 3 is a multi-layer perceptron composed of eight hidden layers and eight output layers. The multilayer perceptron structure is referred to as a fully-connected neural network. In a fully connected neural network, a connection relationship does not exist between nodes located in the same layer, and a connection relationship exists only between nodes located in adjacent layers. DNN has a fully connected neural network structure and is composed of a combination of a number of hidden layers and activation functions, so it can be usefully applied to figure out the correlation between input and output. Here, the correlation characteristic may mean a joint probability of input/output.

On the other hand, depending on how a plurality of perceptrons are connected to each other, various artificial neural network structures different from the above-described DNN can be formed.

4 shows an example of a convolutional neural network to which the method proposed in this specification can be applied, and FIG. 5 shows a filter operation in a convolutional neural network to which the method proposed in this specification can be applied.

In DNN, nodes located inside one layer are arranged in a one-dimensional vertical direction. However, in FIG. 4 , it can be assumed that the nodes are two-dimensionally arranged with w horizontally and h vertical nodes (convolutional neural network structure of FIG. 4 ). In this case, since a weight is added per connection in the connection process from one input node to the hidden layer, a total of h×w weights must be considered. Since there are h×w nodes in the input layer, a total of h ² w ² weights are needed between two adjacent layers.

The convolutional neural network of FIG. 4 has a problem in that the number of weights increases exponentially according to the number of connections, so instead of considering the connection of all modes between adjacent layers, it is assumed that a filter with a small size exists in FIG. As in Fig., the weighted sum and activation function calculations are performed on the overlapping filters.

One filter has a weight corresponding to the number corresponding to its size, and weight learning can be performed so that a specific feature on an image can be extracted and output as a factor. In FIG. 5 , a 3×3 filter is applied to the upper left 3×3 region of the input layer, and an output value obtained by performing weighted sum and activation function operations on the corresponding node is stored in z ₂₂ .

The filter performs weight sum and activation function calculations while moving horizontally and vertically at regular intervals while scanning the input layer, and places the output value at the current filter position. This calculation method is similar to a convolution operation on an image in the field of computer vision, so a deep neural network with such a structure is called a convolutional neural network (CNN), and a hidden layer generated as a result of the convolution operation is called a convolutional layer. Also, a neural network having a plurality of convolutional layers is called a deep convolutional neural network (DCNN).

In the convolution layer, the number of weights can be reduced by calculating the weighted sum by including only nodes located in the region covered by the filter in the node where the filter is currently located. Due to this, one filter can be used to focus on features for the local area. Accordingly, CNN can be effectively applied to image data processing in which physical distance in a two-dimensional domain is an important criterion. Meanwhile, in CNN, a plurality of filters may be applied immediately before the convolution layer, and a plurality of output results may be generated through a convolution operation of each filter.

Meanwhile, there may be data whose sequence characteristics are important according to data properties. Considering the length variability and precedence relationship of the sequence data, one element in the data sequence is input at each timestep, and the output vector (hidden vector) of the hidden layer output at a specific time is input together with the next element in the sequence. A structure in which this method is applied to an artificial neural network is called a recurrent neural network structure.

6 shows a neural network structure in which a cyclic loop to which the method proposed in the present specification can be applied.

Referring to FIG. 6 , a recurrent neural network (RNN) completely connects elements (x ₁ (t), x ₂ (t), ... , x _d (t)) of any gaze t on a data sequence. In the process of input to the neural network, the immediately preceding time point t-1 is weighted by inputting the hidden vectors (z ₁ (t-1), z ₂ (t-1), ... , z _H (t-1)) together. It is a structure to which sum and activation functions are applied. The reason why the hidden vector is transferred to the next time in this way is because it is considered that information in the input vector at the previous time is accumulated in the hidden vector of the current time.

7 shows an operation structure of a recurrent neural network to which the method proposed in the present specification can be applied.

Referring to FIG. 7 , the recurrent neural network operates in a predetermined time sequence with respect to an input data sequence.

The hidden vector (z ₁ (1), z ₂ (1) when the input vector (x ₁ (t), x ₂ (t), ... , x _d (t)) at time 1 is input to the recurrent neural network ), ... , z _H (1)) are input together with the input vectors of time 2 (x ₁ (2), x ₂ (2), ... , x _d (2)) to form weighted sum and activation functions to determine the vectors of the hidden layer (z ₁ (2), z ₂ (2), ... , z _H (2)). This process is repeatedly performed until time point 2, time point 3, ..., and time point T.

Meanwhile, when a plurality of hidden layers are disposed in a recurrent neural network, this is called a deep recurrent neural network (DRNN). The recurrent neural network is designed to be usefully applied to sequence data (eg, natural language processing).

As a neural network core used as a learning method, in addition to DNN, CNN, and RNN, Restricted Boltzmann Machine (RBM), deep belief networks (DBN), Deep Q-Network and It includes various deep learning techniques such as computer vision, speech recognition, natural language processing, voice/signal processing, and can be applied to fields such as derivation of optimal values through large-capacity data processing.

8 is an example of a DNN model to which the method proposed in the present specification can be applied.

A deep neural network (DNN) is an artificial neural network (ANN) composed of several hidden layers between an input layer and an output layer. Deep neural networks can model complex non-linear relationships like general artificial neural networks.

For example, in a deep neural network structure for an object identification model, each object may be expressed as a hierarchical configuration of image basic elements. In this case, the additional layers may aggregate features of the gradually gathered lower layers. This feature of deep neural networks enables modeling of complex data with fewer units (units, nodes) compared to similarly performed artificial neural networks.

As the number of hidden layers increases, the artificial neural network is called 'deep', and the machine learning paradigm that uses the sufficiently deep artificial neural network as a learning model is called deep learning. And, a sufficiently deep artificial neural network used for such deep learning is collectively called a deep neural network (DNN).

In the present invention, data required for training the POI data generation model may be input to the input layer of the DNN, and meaningful data that a user can use may be generated through the output layer while passing through the hidden layers.

In the specification of the present invention, an artificial neural network used for such a deep learning method is collectively referred to as a DNN, but if meaningful data can be output in a similar way, of course, other deep learning methods may be applied.

9 is a block diagram of an AI device according to an embodiment of the present invention.

The AI device 20 may include an electronic device including an AI module capable of performing AI processing, or a server including the AI module. In addition, the AI device 20 is at least a part of the vehicle-infrastructure optical camera communication system 1100 shown in FIG. 11 or the transmission unit recognition device 1200 of the vehicle-infrastructure optical camera communication shown in FIG. 12 . It may be included to perform at least a part of AI processing together.

The AI processing of the AI device 20 is controlled by the transmission unit recognition device 1200 in the vehicle-infrastructure optical camera communication system 1100 shown in FIG. 11 and the vehicle-infrastructure optical camera communication shown in FIG. 12 . It may include all operations related to , and all operations for optical camera communication between vehicle-infrastructure through artificial intelligence learning. For example, the AI device 20 may perform an operation of recognizing a transmitter on a road infrastructure after processing/determining and learning the collected data set by AI processing based on the YOLO model. The AI device 20 may be included as a part of the image processing unit 1122 of FIG. 11 or may be replaced by the image processing unit 1122 . Also, the AI device 20 may be included as a part of the transmitter recognizing device 1200 of FIG. 12 or replaced with the transmitter recognizing device 1200 .

The AI device 20 may include an AI processor 21 , a memory 25 and/or a communication unit 27 .

The AI device 20 is a computing device capable of learning a neural network, and may be implemented with various electronic devices such as a server, a desktop PC, a notebook PC, a tablet PC, or the like, or implemented as a single chip. In the present invention, the AI device 20 may be a vehicle-infrastructure optical camera communication device implemented in the form of any one of the various electronic devices.

The AI processor 21 may learn the neural network using a program stored in the memory 25 . In particular, the AI processor 21 may learn a neural network for recognizing device-related data. Here, a neural network for recognizing device-related data may be designed to simulate a human brain structure on a computer, and may include a plurality of network nodes having weights that simulate neurons of a human neural network. The plurality of network modes may transmit and receive data according to a connection relationship, respectively, so as to simulate a synaptic activity of a neuron in which a neuron sends and receives a signal through a synapse. Here, the neural network may include a deep learning model developed from a neural network model. In a deep learning model, a plurality of network nodes can exchange data according to a convolutional connection relationship while being located in different layers. Examples of neural network models include deep neural networks (DNN), convolutional deep neural networks (CNN), recurrent neural networks (RNN), Restricted Boltzmann Machine (RBM), deep trust It includes various deep learning techniques such as neural networks (DBN, deep belief networks) and deep Q-networks, and can be applied to fields such as computer vision (CV), voice recognition, natural language processing, and voice/signal processing. can

Meanwhile, the processor performing the above-described function may be a general-purpose processor (eg, CPU), but may be an AI-only processor (eg, GPU) for artificial intelligence learning.

The memory 25 may store various programs and data necessary for the operation of the AI device 20 . The memory 25 may be implemented as a non-volatile memory, a volatile memory, a flash-memory, a hard disk drive (HDD), or a solid state drive (SDD). The memory 25 is accessed by the AI processor 21 , and reading/writing/modification/deletion/update of data by the AI processor 21 may be performed. In addition, the memory 25 may store a neural network model (eg, the deep learning model 26 ) generated through a learning algorithm for data classification/recognition according to an embodiment of the present invention.

Meanwhile, the AI processor 21 may include a data learning unit 22 that learns a neural network for data classification/recognition. The data learning unit 22 may learn a criterion regarding which training data to use to determine data classification/recognition and how to classify and recognize data using the training data. The data learning unit 22 may learn the deep learning model by acquiring learning data to be used for learning and applying the acquired learning data to the deep learning model.

The data learning unit 22 may be manufactured in the form of at least one hardware chip and mounted on the AI device 20 . For example, the data learning unit 22 may be manufactured in the form of a dedicated hardware chip for artificial intelligence (AI), or may be manufactured as a part of a general-purpose processor (CPU) or graphics-only processor (GPU) to the AI device 20 . may be mounted. In addition, the data learning unit 22 may be implemented as a software module. When implemented as a software module (or a program module including instructions), the software module may be stored in a computer-readable non-transitory computer readable medium. In this case, the at least one software module may be provided by an operating system (Operating System) or may be provided by an application (application program).

The data learning unit 22 may include a training data acquiring unit 23 and a model learning unit 24 .

The training data acquisition unit 23 may acquire training data required for a neural network model for classifying and recognizing data. For example, the training data acquisition unit 23 may acquire image data and/or sample data for transmission units on a road infrastructure to be input to the neural network model as training data.

The model learning unit 24 may use the acquired training data to learn the neural network model to have a criterion for determining how to classify predetermined data. In this case, the model learning unit 24 may train the neural network model through supervised learning using at least a portion of the training data as a criterion for determination. Alternatively, the model learning unit 24 may learn the neural network model through unsupervised learning for discovering a judgment criterion by self-learning using learning data without guidance. Also, the model learning unit 24 may train the neural network model through reinforcement learning using feedback on whether a result of situation determination according to learning is correct. Also, the model learning unit 24 may train the neural network model by using a learning algorithm including an error back-propagation method or a gradient decent method.

When the neural network model is trained, the model learning unit 24 may store the learned neural network model in a memory. The model learning unit 24 may store the learned neural network model in the memory of the server connected to the AI device 20 through a wired or wireless network.

The data learning unit 22 further includes a training data preprocessing unit (not shown) and a training data selection unit (not shown) in order to improve the analysis result of the recognition model or to save resources or time required for generating the recognition model. You may.

The learning data preprocessor may preprocess the acquired data so that the acquired data can be used for learning for situation determination. For example, the training data preprocessor may process the acquired data into a preset format so that the model learning unit 24 may use the acquired training data for learning for image data recognition for the transmitter.

In addition, the learning data selection unit may select data necessary for learning from among the learning data acquired by the learning data acquiring unit 23 or the training data pre-processed by the preprocessing unit. The selected training data may be provided to the model learning unit 24 . For example, the training data selector may select only data included in the specific field as training data by recognizing a specific field among data sets collected through the network.

In addition, the data learning unit 22 may further include a model evaluation unit (not shown) in order to improve the analysis result of the neural network model.

The model evaluator may input evaluation data to the neural network model and, when an analysis result output from the evaluation data does not satisfy a predetermined criterion, may cause the model learning unit 22 to learn again. In this case, the evaluation data may be predefined data for evaluating the recognition model. As an example, the model evaluator may evaluate as not satisfying a predetermined criterion when, among the analysis results of the learned recognition model for the evaluation data, the number or ratio of evaluation data for which the analysis result is not accurate exceeds a preset threshold. .

The communication unit 27 may transmit the AI processing result by the AI processor 21 to an external electronic device. Here, the external electronic device may be defined as a vehicle-infrastructure optical camera communication system ( FIGS. 11 and 1100 ). Meanwhile, the AI device 20 may be implemented by being functionally embedded in the image processing unit ( FIGS. 11 and 1122 ) provided in the vehicle-infrastructure optical camera communication system.

On the other hand, the AI device 20 shown in FIG. 9 has been functionally divided into the AI processor 21 , the memory 25 , the communication unit 27 , and the like, but the above-described components are integrated into one module and the AI module Note that it may also be called

Hereinafter, an example of a You Only Look Once (YOLO) model to which the method proposed in the present specification can be applied will be described.

The YOLO model used in the machine learning field or deep learning field is a technology that detects a semantic region in an image in real time. It's a pattern. The YOLO model is a supervised learning-type machine learning algorithm, in which the semantic domain to be recognized and its classification information (ie, labeling) are paired to generate training data, and a learning model using thousands of training data as input parameters. It may be configured to include a data generation step and an object classification step of extracting a learned semantic region from an arbitrary image using the learning model data.

As an example of a field that uses the YOLO model, it consists of a pair of vehicle images and classification information (eg, trucks, buses, passenger cars, etc.) If you prepare thousands of training data, run the YOLO model machine learning function using the training data as input parameters to generate model data, and execute the classification function of the YOLO model using the generated model data and arbitrary images as input parameters. The extraction result of the semantic region may be provided in the form of a bounding box and classification information on the bounding box. The YOLO model can be used in real-time object recognition fields such as self-driving cars and person detection in images because of its high performance and accuracy and fast execution speed in the field of machine learning image recognition.

In the object classification step of the YOLO model, the target reliability of the initial detection object is specifically set for accurate detection of the semantic region, and the input image is divided into bounding box units of a certain size. The object detection process calculates the detection candidate object reliability that can be included in the bounding box for all the divided bounding boxes, and extracts the corresponding bounding box when the calculated detection candidate object object reliability is equal to or greater than the initial detection target reliability. If the bounding box is not extracted through the object detection process, the size of the bounding box is adjusted and the object detection process is repeatedly executed.

In addition, the method proposed in the present specification improves the object detection speed and reliability of the YOLO model through transfer learning. Transfer learning is a means to compensate for the lack of learning data in the field of machine learning. It is a technique that transfers knowledge of the source domain to the target domain to train the model in the context of little learning data.

In the application method of transfer learning, the network is pre-trained through data in the source domain first. Afterwards, to apply transfer learning, an additional layer is added to the source domain network to construct a transfer learning network. After that, the transfer learning network is trained using the target domain data to obtain an output for the new target domain. This series of processes is called transfer learning process, and adding a new layer to the pre-trained source domain network to train the transfer learning model in the target domain is called fine-tuning. Transfer learning has the characteristic of showing good performance even with a small amount of data as the knowledge of the pre-trained network is transferred to a new target domain.

Hereinafter, a vehicle-to-infrastructure optical camera communication technology to which the method proposed in the present specification can be applied will be schematically described with reference to the accompanying FIGS. 10 to 15 .

The transmitter recognition method in the vehicle-infrastructure optical camera communication method proposed in the present specification uses a digital camera such as a black box installed in a vehicle or pre-photographed image data to distinguish the types of transmitters installed on the road. Prepare thousands or more of training data consisting of a pair of classification information for images and images of various transmitters, for example, classification information of LED transmitters on road infrastructure such as traffic lights, traffic signs, and electric signs for traffic guidance. Executes the YOLO model machine learning function as an input parameter to generate model data, and executes the classification function of the YOLO model using the generated model data and arbitrary images as input parameters. It can be provided in the form of classification information for In addition to this, in this specification, when setting the bounding box to the extraction result of the semantic region, the region of interest is set only for the LED array that actually outputs light in the image of the entire transmitter, that is, the light source in the vehicle-infrastructure optical camera communication device. This allows the signal reception rate to be further increased.

10 shows the concept of vehicle-infrastructure optical camera communication, and FIG. 11 shows the concept of a vehicle-infrastructure optical camera communication system based on YOLO detection according to an embodiment of the present invention.

Referring to FIG. 10, the optical camera communication between the vehicle and the infrastructure includes an optical camera (not shown) installed in the vehicles 1000 and 1010 such as a traffic light 1020, a traffic sign 1030, a traffic guide electronic sign 1040, etc. Receiving data modulated into optical signals from a transmitter configured including an LED array on road infrastructure. The data modulated by the optical signal may include information about the environment around the vehicle, traffic situation information, and the like, which is difficult to determine only with the sensor of the vehicle. The traffic situation information and the like may be collected, generated, and transmitted by a central control center that manages roads.

In this case, the vehicles 1000 and 1010 photograph the transmitters on the road infrastructure through a camera provided in the vehicle while moving. Since the tracking of the data becomes unstable, the data reception rate is lowered. Therefore, the present invention can improve the recognition rate of the transmitter by executing the learning by the YOLO model in detecting the transmitters on the road infrastructure.

Referring to FIG. 11 , the vehicle-infrastructure optical camera communication system 1100 includes a transmission device 1110 installed in road infrastructure to transmit optically modulated data, and a reception device installed in a vehicle to receive the optically modulated data. 1120 may be included.

The transmitter 1110 may include an encoder 1112 , a modulator 1113 , and a traffic light 1114 .

The encoder 1112 encodes the input data 1111 into a form for transmission, the modulator 1113 modulates the encoded input data to transmit an optical signal, and the traffic light 1114 optically transmits the modulated data. It generates and outputs a signal. In this embodiment, a traffic light 1114 including an LED array is exemplified as an example of a transmitter for outputting an optical signal, but as a transmitter, the traffic sign 1030 of FIG. Any device capable of transmitting a signal can be used.

The reception device 1120 may include a camera 1121 , an image processing unit 1122 , and a decoder 1123 .

The camera 1121 may photograph the surrounding environment including the road infrastructure within the photographing range in order to photograph the traffic light 1114 of the transmitting device 1110 .

The image processing unit 1122 may process the photographed image of the road infrastructure as a YOLO model to recognize the traffic light 1114 and then receive an optical signal.

The decoder 1123 decodes the received optical signal and outputs 1124 the transmitted data.

The illustrated components are not essential, so a vehicle-infrastructure optical camera communication system having more or fewer components may be implemented. These components may be implemented as hardware or software, or may be implemented through a combination of hardware and software.

Meanwhile, a method for the image processing unit 1122 to receive an optical signal using the YOLO model may be performed through the following process.

First, the image processing unit 1122 processes the captured image of the road infrastructure as a YOLO model (S1131).

Next, the image processing unit 1122 detects the traffic light 1114 from the photographed image of the road infrastructure, and then sets a bounding box to the detected traffic light 1114 ( S1132 ).

Next, the image processing unit 1122 outputs a bounding box while continuously tracking the traffic light 1114 in the image data input for data reception (S1133).

Finally, the image processing unit 1122 extracts the data by extracting the optical signal from the image of the traffic light 1114 in the output bounding box (S1134).

On the other hand, in the above-described transmitting device 1110, all other components except the traffic light 1114 may be installed in a separate place (eg, central control center) different from the place where the traffic light 1114 is installed, The traffic light 1114 and the central control center may be connected to a separate network to transmit data obtained by encoding and modulating input data to be transmitted from the central control center to the traffic light 1114 through the network.

The networks disclosed herein include, for example, wireless networks, wired networks, public networks such as the Internet, private networks, Global System for Mobile communication network (GSM) networks, and general packet wireless networks (General). Packet Radio Network (GPRN), Local Area Network (LAN), Wide Area Network (WAN), Metropolitan Area Network (MAN), Cellular Network, Public Switched Telephone Network ; PSTN), Personal Area Network, Bluetooth, Wi-Fi Direct, Near Field communication, Ultra-Wide band, a combination thereof, or any may be, but not limited to, other networks of

12 illustrates a transmitter recognition apparatus 1200 in vehicle-infrastructure-to-infrastructure optical camera communication according to an embodiment of the present invention. The transmitter recognition apparatus 1200 may include all functions of the image processing unit 1122 of FIG. 11 .

12 , the apparatus 1200 for recognizing a transmitter in vehicle-infrastructure optical camera communication according to an embodiment of the present invention includes a data receiver 1210 , a training data generator 1220 , a model learner 1230 , and a transmitter recognition unit 1240 .

The data receiver 1210 may receive image data for a transmitter existing on the road infrastructure photographed by the camera. The transmitter on the road infrastructure may be configured to include an LED array such as a traffic light, a traffic sign, and an electric sign for traffic guidance.

The training data generator 1220 may configure a data set by matching a bounding box for a transmitter on the road infrastructure and classification information for the transmitter. In particular, the learning data generating unit 1220 may set the bounding box only for the light source part of the transmitter. That is, the bounding box may be set to surround only the LED array that outputs the optical signal from the transmitter, not the entire transmitter. The bounding box may be set to be slightly larger than the LED array to include the LED array in the bounding box, but may be set to remove unnecessary parts other than the LED array as much as possible.

Also, the training data generator 1220 may generate a data set in which a weight for detecting a transmitter is updated by labeling the transmitter in image data for infrastructure on the road.

The model learning unit 1230 may train the YOLO model using the data set.

The transmitter recognition unit 1240 may recognize the bounding box output by processing the image data for the transmitter on the road infrastructure in the YOLO model learned through the model learning unit 1230 as a transmitter of optical camera communication. The transmitter recognition unit may extract an optical camera communication signal from an image of a light source in a bounding box.

13 shows an example of transmitter recognition, and shows an example of a bounding box output result for the transmitter by the YOLO model. Referring to FIG. 13, the YOLO model shows a bounding box 1310 surrounding the entire transmitter. It is set and output, and information 1320 about the classification of the transmitter in the corresponding bounding box is also output. However, when the bounding box 1310 surrounding the entire transmitter is used, the area actually occupied by the LED in the bounding box 1310 becomes a small part, so the difference in the average intensity of light in the bounding box for the On or Off state of the LED Since is not large, it may be difficult to detect a received signal.

However, as shown in FIG. 14 , the apparatus for recognizing the transmitter according to an embodiment of the present invention processes the image including the road infrastructure by the YOLO model, and then the transmitter includes only the light source in the bounding box 1410 and 1420. can be set and output, since the area actually occupied by the LED in the bounding boxes 1410 and 1420 occupies a large proportion, the difference in average intensity of light in the bounding box with respect to the On or Off state of the LED becomes large. Therefore, the apparatus for recognizing a transmitter according to an embodiment of the present invention can easily detect a received signal.

14 shows an example of the recognition result of the transmitter by the YOLO model in the apparatus for recognizing the transmitter according to an embodiment of the present invention, and FIG. 14 (a) is a bounding box 1410 for the transmitter when the transmitter is a traffic light An example of the output result is shown, and FIG. 14(b) shows an example of the output result of the bounding box 1420 for the transmitter when the transmitter is a street lamp.

Optical communication data can be received by separately extracting only the area where the LED is detected through intensity histogram analysis of the light output from the bounding box. Delays may occur.

15 illustrates a transmitter detection method of the transmitter recognizing apparatus 1200 according to an embodiment of the present invention.

Referring to FIG. 15 , the transmitter recognition method performed by the apparatus for recognizing a transmitter of vehicle-infrastructure optical camera communication according to an embodiment of the present invention includes receiving image data for a transmitter on a road infrastructure (S1510), road infrastructure A data set is configured by matching a bounding box for the transmitter of the image and classification information about the transmitter, but setting a bounding box only for the light source part in the transmitter (S1520), a YOLO model using the data set It may include learning (S1530), and recognizing a bounding box output by processing image data for road infrastructure input to the learned YOLO model as a transmitter of optical camera communication (S1540).

Here, the configuring of the data set may include updating the weight for detecting the transmitter by labeling the transmitter in the image data. In the data set, the transmitter may be labeled for each type. In addition, the transmitter includes an LED traffic light, an LED street light, an LED traffic information sign, an LED traffic guide sign, and an LED advertising board. It is possible.

In addition, the step of setting the bounding box may include setting a region including only the LED light source portion in the entire image for the transmitter as the bounding box.

In addition, the transmission unit recognition method performed by the apparatus for recognizing the transmission unit of vehicle-infrastructure optical camera communication according to an embodiment of the present invention may further include extracting an optical camera communication signal from a light source in the bounding box.

Meanwhile, according to another embodiment of the present invention, the transmitter recognition method performed by the transmitter recognition device for vehicle-infrastructure optical camera communication receives image data for the transmitter on the road infrastructure and selects a data set for learning the YOLO model. constructing; training a YOLO model using the data set; and processing the image data input to the learned YOLO model and outputting a recognition result for the transmitter on the road infrastructure, but setting and outputting a bounding box only for the light source part in the entire image of the recognized transmitter. have.

In the above description, the steps may be further divided into additional steps or combined into fewer steps, depending on the embodiment of the present invention. In addition, some steps may be omitted as necessary, and the order between the steps may be switched.

As used herein, the term “unit” (eg, a control unit, etc.) may mean a unit including, for example, one or a combination of two or more of hardware, software, or firmware. The term “unit” may be used interchangeably with terms such as, for example, unit, logic, logical block, component, or circuit. A "part" may be a minimum unit of a component integrally formed or a part thereof. A “unit” may be a minimum unit or a part of performing one or more functions. “Part” may be implemented mechanically or electronically. For example, a “unit” may be one of an Application-Specific Integrated Circuit (ASIC) chip, Field-Programmable Gate Arrays (FPGAs), or programmable-logic device that performs certain operations, known or to be developed in the future. It may include at least one.

At least a portion of an apparatus (eg, modules or functions thereof) or a method (eg, operations) according to various embodiments is, for example, a computer-readable storage medium in the form of a program module It can be implemented as a command stored in . When the instruction is executed by a processor, the one or more processors may perform a function corresponding to the instruction. The computer-readable storage medium may be, for example, a memory.

Computer-readable storage media/computer-readable recording media include hard disks, floppy disks, magnetic media (eg, magnetic tape), and optical media (eg, compact CD-ROMs). disc read only memory), digital versatile disc (DVD), magneto-optical media (e.g., floptical disk), hardware devices (e.g., read only memory (ROM), random (RAM) access memory), or flash memory, etc.) In addition, program instructions may include high-level language code that can be executed by a computer using an interpreter, etc. as well as machine code such as generated by a compiler. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the various embodiments, and vice versa.

A module or program module according to various embodiments may include at least one or more of the above-described components, some may be omitted, or may further include additional other components. Operations performed by a module, a program module, or other components according to various embodiments may be executed sequentially, in parallel, iteratively, or in a heuristic manner. Also, some operations may be executed in a different order, omitted, or other operations may be added.

As used herein, the term “a” is defined as one or more than one. Also, the use of introductory phrases such as “at least one” and “one or more” in a claim includes introductory phrases such as “at least one” and “one or more” and an obscure phrase such as “a” in the same claim. The introduction of another claim element by the ambiguous phrase "an", even if there is any, shall be construed to mean to limit any particular claim containing the claim element so introduced, to an invention containing only one such element. shouldn't be

Unless otherwise specified, terms such as “first” and “second” are used to arbitrarily distinguish the elements described by such terms. Accordingly, these terms are not necessarily intended to indicate a temporal or other priority of such elements, and the mere fact that certain measures are recited in different claims does not indicate that a combination of these measures cannot be used to advantage. . Accordingly, these terms are not necessarily intended to indicate the temporal or other priorities of such elements. The mere fact that certain measures are cited in different claims does not indicate that a combination of these measures cannot be useful.

The arrangement of components to achieve the same function is effectively “related” such that the desired function is achieved. Thus, any two components combined to achieve a particular functionality may be considered "related" to each other so that the desired function is achieved, regardless of structure or intervening component. Likewise, two components thus associated may be considered “operably connected” or “operably coupled” to each other to achieve a desired function.

Also, those skilled in the art will recognize that the boundaries between the functionality of the operations described above are illustrative only. A plurality of operations may be combined into a single operation, the single operation may be distributed into additional operations, and the operations may be executed at least partially overlapping in time. Also, alternative embodiments may include multiple instances of a particular operation, and the order of operations may be changed in various other embodiments. However, other modifications, variations and alternatives are also possible. Accordingly, the detailed description and drawings are to be regarded in an illustrative rather than a restrictive sense.

The phrase “may be X” indicates that condition X may be satisfied. This phrase also indicates that condition X may not be met. For example, a reference to a system that includes a particular component should also include a scenario in which the system does not contain the specific component. For example, a reference to a method that includes a particular action should also include a scenario in which the method does not include the particular component. As another example, however, references to a system configured to perform a specific action should also include scenarios in which the system is not configured to perform a specific action.

The terms “comprising,” “having,” “consisting of,” “consisting of, and “consisting essentially of are used interchangeably. For example, any method may include at least the acts contained in the drawings and/or the specification, and may include only the acts contained in the drawings and/or the specification.

One of ordinary skill in the art would recognize that the boundaries between logical blocks are exemplary only, and that alternative embodiments may incorporate logical blocks or circuit elements or impose an alternative decomposition of functionality on the various logical blocks or circuit elements. will recognize that Accordingly, it should be understood that the architecture shown herein is exemplary only, and in fact, many other architectures may be implemented that achieve the same functionality.

Also, for example, in one embodiment, the illustrated examples may be implemented as circuitry located on a single integrated circuit or within the same device. Alternatively, the above examples may be implemented as any number of individual integrated circuits or individual devices interconnected to each other in any suitable manner, and other variations, modifications, variations and alternatives are also possible. Accordingly, the specification and drawings are to be regarded in an illustrative rather than a restrictive sense.

Also, for example, the examples or portions thereof may be implemented as a software or code representation of a physical circuit or a logical representation convertible into a physical circuit, such as any suitable type of hardware description language.

Further, although the present invention is not limited to physical devices or units implemented in non-programmable hardware, mainframes, minicomputers, servers, workstations, personal computers, notepads, generally referred to herein as 'computer systems' , personal digital assistants (PDAs), electronic games, automobiles and other embedded systems, mobile phones and various other wireless devices, such as programmable devices capable of performing desired device functions by operating in accordance with appropriate program code. Alternatively, it may be applied to units as well.

A system, apparatus, or device referred to in this specification includes at least one hardware component.

Connections as described herein may be any type of connection suitable for transmitting a signal to or from each node, unit or apparatus, for example via an intermediate apparatus. Thus, unless implied or otherwise stated, a connection may be, for example, a direct connection or an indirect connection. A connection may be described or depicted with reference to a single connection, multiple connections, one-way connection, or two-way connection. However, different embodiments may change the implementation of the connection. For example, you can use a separate one-way connection rather than a two-way connection, and vice versa. In addition, a plurality of connections may be replaced with a single connection that transmits a plurality of signals sequentially or in a time multiplexed manner. Likewise, a single connection carrying multiple signals may be split into various connections carrying subsets of these signals. Therefore, many options exist for transmitting the signal.

One of ordinary skill in the art would recognize that the boundaries between logical blocks are exemplary only, and that alternative embodiments may incorporate logical blocks or circuit elements or impose an alternative decomposition of functionality on the various logical blocks or circuit elements. will recognize that Accordingly, it should be understood that the architecture shown herein is exemplary only, and in fact, many other architectures may be implemented that achieve the same functionality.

In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word 'comprising' does not exclude the presence of elements or acts listed in a claim.

In the above, a preferred embodiment of the technology of the present specification has been described with reference to the accompanying drawings. Here, the terms or words used in the present specification and claims should not be construed as being limited to conventional or dictionary meanings, but should be interpreted as meanings and concepts consistent with the technical spirit of the present invention. The scope of the present invention is not limited to the embodiments disclosed herein, and the present invention may be modified, changed, or improved in various forms within the scope of the spirit and claims of the present invention.

20: AI device 21: AI processor
22: data learning unit 23: learning data acquisition unit
24: model learning unit 25: memory
26: neural network model 27: communication department
1000, 1010: vehicle 1020: traffic light
1030: traffic sign 1040: traffic information electronic sign
1100: vehicle-infrastructure optical camera communication system
1110: transmitting device 1111: input data
1112: encoder 1113: modulator
1114: traffic light 1120: receiving device
1121: camera 1122: image processing unit
1123: decoder 1124: output data
1200: transmitter recognizing unit 1210: data receiver
1220: training data generation unit 1230: model training unit
1240: transmitter recognition unit

Claims (9)

In the transmission unit recognition method performed by the transmission unit recognition device of the vehicle-infrastructure optical camera communication,
receiving image data for a transmitter on a road infrastructure;
configuring a data set by matching a bounding box for a transmitter on the road infrastructure and classification information for the transmitter, but setting a bounding box only for a light source in the transmitter;
training a YOLO model using the data set; and
recognizing a bounding box output by processing image data for road infrastructure input to the learned YOLO model as a transmitter of optical camera communication;
A method of recognizing a transmitter in vehicle-infrastructure-to-infrastructure optical camera communication comprising a. The method of claim 1,
The step of configuring a data set by matching a bounding box for the transmitter on the road infrastructure and classification information for the transmitter, setting a bounding box only for the light source part in the transmitter,
updating a weight for detecting a transmitter by labeling the transmitter in the image data;
A method for recognizing a transmitter in vehicle-infrastructure-to-infrastructure optical camera communication, comprising: The method of claim 1,
The step of configuring a data set by matching a bounding box for the transmitter on the road infrastructure and classification information for the transmitter, setting a bounding box only for the light source part in the transmitter,
setting an area including only the LED light source part in the entire image for the transmitter as a bounding box;
A method for recognizing a transmitter in vehicle-infrastructure-to-infrastructure optical camera communication, comprising: 4. The method of claim 3,
extracting an optical camera communication signal from a light source in the bounding box;
Vehicle-to-infrastructure optical camera communication, characterized in that it further comprises a transmission unit recognition method. a data receiver for receiving image data for a transmitter on a road infrastructure;
a learning data generator configured to configure a data set by matching a bounding box for a transmitter on the road infrastructure and classification information for the transmitter, but set a bounding box only for a light source in the transmitter;
a model learning unit for learning the YOLO model using the data set; and
a transmitter recognition unit for recognizing a bounding box output by processing image data for a transmitter on the road infrastructure in the learned YOLO model as a transmitter for optical camera communication;
A vehicle-to-infrastructure-to-infrastructure optical camera communication transmitter recognition device comprising a. The method of claim 5, wherein the training data generator
Transmitter recognition apparatus in vehicle-infrastructure optical camera communication, characterized in that by labeling the transmitter in the image data to update a weight for detecting the transmitter. The method of claim 5, wherein the training data generator
Transmitter recognition device in vehicle-infrastructure optical camera communication, characterized in that the region including only the LED light source part in the entire image for the transmitter is set as a bounding box. The method of claim 7, wherein the transmitter recognition unit
Transmitter recognition device in vehicle-infrastructure optical camera communication, characterized in that extracting the optical camera communication signal from the light source in the bounding box. In the transmission unit recognition method performed in the transmission unit recognition device of the vehicle-infrastructure optical camera communication,
composing a data set for training a YOLO model by receiving image data for a transmitter on a road infrastructure;
training a YOLO model using the data set; and
processing the image data input to the learned YOLO model and outputting a recognition result for the transmitter on the road infrastructure, setting and outputting a bounding box only for the light source part in the entire image of the recognized transmitter;
A method of recognizing a transmitter in vehicle-infrastructure-to-infrastructure optical camera communication comprising a.

Images