TWI591593B - Traffic signs, traffic lights and timers for video detection and identification of voice announcers - Google Patents

Description

Voice broadcaster for traffic signs, traffic lights and image recognition and identification of the reader

The invention relates to a voice broadcaster for traffic sign, traffic light number and image recognition and identification of a countdown device, in particular to use an electronic image recognition technology to assist in identifying a traffic sign, a light number or a countdown number in front, and broadcast it by voice. The method informs the driver of the electronic device, so that the driver does not need to distract the traffic signs on the road to diversify his attention and improve driving safety.

With the advancement of today's computer technology, human life has gradually changed from the era of manual work to the era of automated computer. In the development of modern transportation vehicles, the traffic conditions are becoming more and more complicated. It is an important issue to use the sensing system to provide the surrounding environment and related information to the driver to increase the safety and effectiveness of driving.遂 “Road Marking Information Prompt Device” (Republic of China Invention Patent No. M250157), “Wireless Traffic Sign Warning System” (Republic of China Invention Patent No. I312492) and “Road Image Identification Navigation Device and Navigation Method” (Republic of China Techniques such as invention patent No. I359259 have been developed. In addition, in order to accelerate the goal of "smart car", many autopilots studied by researchers now need to use related technologies to enable them to safely drive on the road. However, traffic signs are still an important way for current road drivers, due to the size and appearance of traffic signs. Color and content have certain regulations. For example, the general warning sign must use the regular triangle mark, white is used for the background color, pattern or text of the mark, red is used for warning or prohibition sign, etc., plus the type of traffic sign. Very numerous, such as triangular signs, round signs, rectangular signs, diamond signs, etc., the images inside are also very strange, in addition to red, orange, blue, yellow and other various colors, and in the natural environment There will be many factors that adversely detect and identify traffic signs, such as weather (clear, cloudy, rainy), time (early morning, dusk, evening), traffic signs being shaded by trees or cars, traffic signs being deformed, traffic signs There are objects of the same color in the vicinity to detect the interference, consider the driving speed, and quickly identify the validity of the response. It shows that the identification technology has its uniqueness. The relevant identification technology has been widely developed in the current public and private research institutions. However, there is no such thing as the present invention, focusing on the application of relevant traffic sign recognition technology and control technology, and using voice broadcast to avoid road drivers distracting attention from traffic signal interpretation and manipulation, making road drivers easier and safer. The line is on the road.

According to the statistics of the Police Department of the Ministry of the Interior, the number of people who died in traffic accidents in Taiwan every year has remained at around 3,000. The cause of the accident was analyzed. Ninety-six percent were caused by improper driving and improper driving. There are quite a few reasons, such as drunk driving, driver distraction, driver discomfort, etc. Neglect or failure to notice warnings from traffic signs is one of the reasons for improper driving. The driver will not notice the warning from the traffic sign for various reasons. This may result in a car accident or a violation of the traffic sign. A study has found that there are too many traffic signs in places where there are too many traffic signs at the intersection. Leading to driver distraction and causing a car accident, so how to use machine vision technology to detect, identify and reward these traffic signs, to make road driving aware with the least visual burden, in order to improve traffic safety is A new and important topic is the problem that the invention will solve.

The purpose of the present invention is to utilize the latest electronic image and voice recognition technology to detect, identify and select the return traffic number to prevent the road driver from distracting the attention of the traffic sign and increase the safety of the car. That is to say, using the latest traffic sign image recognition technology and the voice or gesture recognition control technology, the road driver can detect and select the driving directions on the road with the least visual burden, and can travel on the road more easily and safely.

The mobile image recognition traffic signal broadcaster capable of achieving the above object is mainly composed of: a photographic lens, a motion processing device, a microphone, a back lens with a 3D image sensing chip, a voice output interface, and a gesture detection control software model. The group, voice recognition voice control software module, traffic sign detection classification and identification software module, traffic light number and countdown detection and identification software module, voice playback software module and function selection output input software interface. Use the latest image traffic sign recognition technology and voice or gesture control technology to prevent road drivers from distracting attention due to the interpretation of traffic signs, allowing road drivers to detect and select driving directions on the road with minimal visual burden To make it easier and safe for road drivers to walk on the road.

The invention provides a voice broadcaster for detecting and recognizing a traffic sign image, comprising: a photographing lens for capturing an image in front of driving; and a traffic sign image detecting and identifying software module for receiving the photographing lens. The image is taken as a classification and identification of the traffic sign image; a mobile computing processing device is used to process and classify the image classified and identified by the traffic sign image detection classification and recognition software module; and a voice playback software module and A voice output interface receives the processing result of the action processing device and selects a corresponding voice file to play.

The mobile computing processing device comprises a high-speed and high-performance multi-core processor combined with a cooperative computing processor, a controller and a memory, and an input/output interface can be connected with the photographic lens and the microphone. It also includes: WiFi communication, Bluetooth communication, 3G communication, 4G communication, GPS receiver and 3D space gravity acceleration detector. The utility model further comprises a microphone and a voice recognition software module for voice recognition. The utility model further comprises a 3D image sensing chip rear lens and a gesture detection control software module, which provides a non-touch gesture control function, and directly activates, selects or turns off the relevant detection and recognition function by using a gesture command. The traffic sign image detection classification and identification software module is used for traffic sign detection according to the color and shape of the flag set in the "Road Traffic Sign Marking Code Setting Rules" published by the government. At different light intensities, the color will be slightly different. In order to make the specific color easy and correct to be extracted, the color filtering method can use the RGB color system to set the RGB value and the scale range to extract the target. Or the range of the hue under the HSI color system is used to extract the target. The traffic sign image detection classification and identification software module is based on the flag setting specification set by the government's published road traffic sign marking code setting rule to give an appropriate template, using the template comparison method or Zernike moments. The distance alignment method or the use of adjacent eight-point directions to find the contour, and the inner contour and the basic circular, triangular, semi-circular and rectangular mathematical rules or SVM method for traffic sign classification. The traffic sign image detection classification and identification software module can perform full or selective identification of the symbols, numbers and texts such as warnings, prohibitions and indications according to the settings. The voice playback software module and the voice output interface in the voice output interface can be a general wired audio jack output and a wireless Bluetooth audio output.

The invention provides a voice broadcaster for traffic signal number and countdown image detection and identification, which comprises: a photographing lens for capturing images in front of driving; a traffic light and reading The second device detects and recognizes the software module, receives the image captured by the photographic lens, classifies and recognizes the traffic signal and the image of the chronograph; and an action processing device that processes the operation via the traffic signal and the chronograph Detecting and recognizing the image classified and recognized by the software module; and a voice playing software module and a voice output interface, receiving the processing result of the motion computing processing device, and selecting the corresponding voice file to broadcast.

The mobile computing processing device comprises a high-speed and high-performance multi-core processor combined with a cooperative computing processor, a controller and a memory, and an input/output interface can be connected with the photographic lens and the microphone. It also includes: WiFi communication, Bluetooth communication, 3G communication, 4G communication, GPS receiver and 3D space gravity acceleration detector. The utility model further comprises a microphone and a voice recognition software module for voice recognition. The utility model further comprises a 3D image sensor film back lens and a gesture detection control software module, which provides a non-touch gesture control function, and directly activates, selects or turns off the relevant detection and recognition function by using a gesture command. The traffic light and the countdown detector detection and identification software module are based on the light number set by the government's published road traffic sign markings and the color and color of the countdown display. Measurement identification. The traffic light and the countdown detection and recognition software module provide a countdown recognition synchronization broadcast function when the countdown display is detected, so that the number of seconds of the countdown display is synchronized with the number of seconds of the voice broadcast. The traffic light and the countdown detection and recognition software module can set the number of seconds to start the countdown broadcast, and promptly drive the voice with a synchronous voice. The voice playback software module and the voice output interface in the voice output interface can be a general wired audio jack output and a wireless Bluetooth audio output.

The invention provides a voice broadcaster for traffic sign, traffic light number and image recognition and identification of a countdown device, which comprises: a photographing lens for capturing an image in front of driving; a traffic The image detection classification and recognition software module receives the image captured by the camera lens and classifies and identifies the traffic sign image; a traffic light number and a countdown device detection and recognition software module receives the camera lens The image is captured and identified by the traffic signal and the image of the reader; a mobile computing device is used to process the classification and identification software module and the traffic signal and the countdown detector. Identifying the image classified and recognized by the software module; and a voice playing software module and a voice output interface, receiving the processing result of the motion computing processing device, and selecting the corresponding voice file to broadcast.

The mobile computing processing device comprises a high-speed and high-performance multi-core processor combined with a cooperative computing processor, a controller and a memory, and an input/output interface can be connected with the photographic lens and the microphone. It also includes: WiFi communication, Bluetooth communication, 3G communication, 4G communication, GPS receiver and 3D space gravity acceleration detector. The utility model further comprises a microphone and a voice recognition software module for voice recognition. The utility model further comprises a 3D image sensor film back lens and a gesture detection control software module, which provides a non-touch gesture control function, and directly activates, selects or turns off the relevant detection and recognition function by using a gesture command. The traffic sign image detection classification and identification software module is used for traffic sign detection according to the color and shape of the flag set in the "Road Traffic Sign Marking Code Setting Rules" published by the government. At different light intensities, the color will be slightly different. In order to make the specific color easy and correct to be extracted, the color filtering method can use the RGB color system to set the RGB value and the scale range to extract the target. Or the range of the hue under the HSI color system is used to extract the target. The traffic sign image detection classification and identification software module is based on the flag setting specification set by the government's published road traffic sign marking code setting rule to give an appropriate template, using the template comparison method or Zernike moments. Distance comparison or use the neighboring eight points to find Contours, classification of traffic signs with inner contours and basic circular, triangular, semi-circular and rectangular mathematical rules or SVM methods. The traffic sign image detection classification and identification software module can perform full or selective identification of the symbols, numbers and texts such as warnings, prohibitions and indications according to the settings. The traffic light and the countdown detector detection and identification software module are based on the light number set by the government's published road traffic sign markings and the color and color of the countdown display. Measurement identification. The traffic light and the countdown detection and recognition software module provide a countdown recognition synchronization broadcast function when the countdown display is detected, so that the number of seconds of the countdown display is synchronized with the number of seconds of the voice broadcast. The traffic light and the countdown detection and recognition software module can set the number of seconds to start the countdown broadcast, and promptly drive the voice with a synchronous voice. The voice playback software module and the voice output interface in the voice output interface can be a general wired audio jack output and a wireless Bluetooth audio output.

The invention has the following advantages when compared with other conventional technologies:

1. The invention can use electronic image recognition technology to assist in identifying the traffic signs, lights or countdown digits in front, and inform the driver by means of voice broadcast, so that the driver does not need to distract the traffic signs on the road when driving. Force to improve driving safety. .

2. The invention can utilize a microphone for voice capture input and a voice recognition voice control software module, so that the driver can start, select or turn off the relevant detection and identification function directly by using the voice sound command without distracting the eye.

3. The invention can utilize a 3D image sensor chip back lens and a gesture detection control software module to provide a non-touch gesture control function, so that driving can be started and selected by gesture instructions without distracting attention. Or turn off the relevant detection and identification function to further improve driving safety.

4. The invention has the function of counting the count recognition, which can synchronize the number of seconds of the countdown display with the number of seconds of the voice broadcast, and can set the number of seconds to start the countdown broadcast, and promptly use the voice to remind the driver to let the driver wait for the red light. Instead of listening to the visual green light, you can make your eyes look a little rest, making it easier and safer to drive traffic lights and other traffic lights.

The detailed description of the preferred embodiments of the present invention is intended to be limited to the scope of the invention, and is not intended to limit the scope of the invention. The patent scope of this case.

In summary, this case is not only innovative in terms of space type, but also can enhance the above-mentioned multiple functions compared with the customary items. It should fully meet the statutory invention patent requirements of novelty and progressiveness, and apply for it according to law. This invention patent application, in order to invent invention, to the sense of virtue.

11‧‧‧Photographic lens

12‧‧‧Mobile computing device

13‧‧‧ microphone

14‧‧‧With 3D image sensing chip and back lens

15‧‧‧Voice output interface

16‧‧‧ gesture detection control software module

17‧‧‧Sound recognition voice control software module

18‧‧‧ Traffic Sign Image Detection Classification and Identification Software Module

19‧‧‧ Traffic Lights and Countdown Detection and Identification Software Module

20‧‧‧Voice playback software module

21‧‧‧ Function selection output input software interface

FIG. 1 is a schematic structural diagram of a voice broadcaster for detecting and recognizing a traffic sign, a traffic light number, and a countdown image of the present invention.

The technical features, contents, and advantages of the present invention, as well as the advantages thereof, can be understood by the reviewing committee, and the present invention will be described in detail with reference to the accompanying drawings. The subject matter is only for the purpose of illustration and supplementary instructions. It is not necessarily the true proportion and precise configuration after the implementation of the invention. Therefore, the proportion and configuration of the attached drawings should not be used. Interpretation of Relationships The limitations of the scope of the invention in actual practice are set forth above.

Please refer to FIG. 1 , which is a schematic structural diagram of a voice broadcaster for detecting and recognizing a traffic sign, a traffic light number, and a countdown image of the present invention. As shown in the figure, a photographic lens 11 , a microphone 13 , and a 3D image are mounted. After the wafer is processed, the back lens 14 and the voice output interface 15 are formed on the mobile processing device 12 to form a hardware body of the voice broadcaster of the traffic sign, the traffic light number and the image recorder of the present invention, and the traffic sign is used as a traffic sign. The detection and identification software module 18 and the traffic signal and the countdown detection and recognition software module 19 are used to detect and recognize the traffic signs and lights captured in the image by the photographic lens 11 and to detect the gestures. The control software module 16 or the voice recognition voice control software module 17 recognizes the input voice signal command or the image gesture command, and selects the traffic signal and the light signal according to the voice input command or the image gesture command by using the function output output software interface 21 The detection and identification function, and then the voice playback software module 20 selects the voice file corresponding to the recognition result, and transmits it to the speaker and the car audio through the voice output interface 15. BlueTooth headsets and other players. In addition to the above main architecture, you can also add a communication port (a wireless transceiver such as WiFi or Bluetooth or 3G or 4G), a GPS receiver and a three-dimensional gravity acceleration detector to provide more integration. Action service features.

The photographic lens 11 is a capture device for the front image during driving, and provides the image detection and classification software module 18 and the traffic signal and the reading and detecting device to detect and recognize the image required by the software module 19 source.

The mobile computing processing device 12 is a high-speed and high-performance multi-core processor, and is input and input channels and a cooperative computing processor, a controller, a memory, a communication port, and a voice output interface 15 to become a hardware input for digital processing. The output device provides the software with the ability to input, calculate, process, transmit, control and output various digital data, and can also be connected to the photographic lens 11, The microphone 13 has a 3D image sensing chip and the back lens 14 and the voice output interface 15 are connected to a voice playing device to form a hardware of the voice broadcaster for the traffic sign, the traffic signal and the image recognition and recognition of the reader.

The microphone 13 provides the voice capture input device of the voice broadcaster for the traffic sign, the traffic light number and the countdown image detection and identification, and provides the voice source of the voice recognition software module 17 for the voice recognition, and the recognized voice signal command It can be used to select the detection and identification function of traffic signs and lights, so that driving can start, select or turn off the relevant detection and identification functions directly without using distracting attention.

The 3D image sensing chip and the back lens 14 are used as a source of gesture image input for the traffic sign, the traffic light number and the read-write image detection and identification of the voice broadcaster, and can provide preliminary 3D gesture recognition, and cooperate with The gesture detection control software module 16 provides a non-touch gesture control function, so that the driver can directly activate, select or turn off the relevant detection and recognition function by using the gesture command to further improve the driving safety without distracting the attention.

The voice output interface 15 is a sound output device for the identification result of the traffic signal, the traffic light number and the readout image detection and identification of the voice broadcaster, when the traffic sign detection classification and identification software module 18 or the traffic signal number And the reading and detecting software module 19, according to the voice signal command or the image gesture command, select the detection and identification function of the traffic sign and the light signal, and after obtaining the relevant identification result, the voice playing software module 20 can select the phase. The voice file corresponding to the identification result is transmitted to the speaker, the car stereo or the BlueTooth earphone through the voice output interface 15.

The gesture detection control software module 16 mainly uses a 3D gesture recognition function provided by the back lens 14 with the 3D image sensing chip, and converts the input gesture image into a function. Select the command and use the output input software interface 21 to select the detection and identification function of the relevant traffic signs and lights.

The voice recognition software module 17 for voice recognition is to identify the voice signal input through the microphone 13 by digital sampling operation, and then find out the corresponding voice command, and then select the corresponding input software interface 21 to select the corresponding voice. Detection and identification of traffic signs and lights.

Among them, the traffic sign detection classification and identification software module 18 is detected in accordance with the color and shape of the flag set in the government-issued road traffic sign marking code setting rule. Under the light intensity, the color will be slightly different. In order to make the specific color easy and correct to be extracted, the color screening method can use the RGB color system to set the RGB value and scale range to extract the target. Or use HSI. The hue of the color system extracts the target. When a possible traffic sign is detected, it can be appropriately given according to the flag setting specification set by the government's published road traffic sign marking rules. Templates using the template comparison method or Zernike moments distance comparison method or using the adjacent eight-point direction to find the contour, and the inner contour and the basic circular, triangular, semi-circular and rectangular mathematical rules or SVM method for traffic sign classification After completing the classification of the traffic signs, the settings of the output input software interface 21 can be selected according to the function to warn, prohibit and indicate Within a logo style, and so do the numbers and text in full or selective identification.

The traffic light number and the countdown detection and identification software module 19 are also based on the light number set by the government's published road traffic sign marking code setting rules and the style and color of the countdown display. Measurement. When the relevant lamp number and its countdown display are detected, in addition to recognizing the lamp number by the general color image recognition method, a countdown display is detected. For digital, the image digital recognition comparison provides the count recognition, and the time and the processing time of the "action image recognition traffic signal broadcaster" itself are used to estimate the time difference between the number of seconds the correction will be displayed and the number of seconds to be broadcast. In order to accurately synchronize the number of seconds of the countdown display with the number of seconds of the voice announcement, and start the number of seconds of the countdown broadcast according to the setting, the precise synchronization of the voice reminds the driving of the countdown seconds currently displayed. Let the driver wait for the red light to replace the visual waiting green light, so that the attention can be a little rest, reduce driving fatigue and increase the safety of the car.

The voice playing software module 20 can select the voice file corresponding to the identification result by using the traffic sign detection classification and identification software module 18 or the traffic light number and the result of the recognition of the reading and detecting software module 19. The voice output interface 15 is delivered to a speaker, a car stereo, or a BlueTooth headset.

The function selection output input software interface 21 is mainly based on the gesture detection control software module 16, and the input gesture image is converted into a function selection command or the voice recognition software module 17 converts the input voice signal into The function selection command is used to select the detection and identification function of the relevant traffic signs and lights.

The detailed description of the present invention is intended to be illustrative of a preferred embodiment of the invention, and is not intended to limit the scope of the invention. The patent scope of this case.

To sum up, this case is not only innovative in terms of technical thinking, but also has many of the above-mentioned functions that are not in the traditional methods of the past. It has fully complied with the statutory invention patent requirements of novelty and progressiveness, and applied for it according to law. Approved this invention patent application, in order to invent invention, to the sense of virtue.

11‧‧‧Photographic lens

12‧‧‧Mobile computing device

13‧‧‧ microphone

14‧‧‧With 3D image sensing chip and back lens

15‧‧‧Voice output interface

16‧‧‧ gesture detection control software module

17‧‧‧Sound recognition voice control software module

18‧‧‧ Traffic Sign Detection Classification and Identification Software Module

19‧‧‧ Traffic Lights and Countdown Detection and Identification Software Module

20‧‧‧Voice playback software module

21‧‧‧ Function selection output input software interface

Claims (8)

A voice broadcaster for traffic signal and countdown image detection and identification, comprising: a photographic lens for capturing an image in front of driving; a traffic light and a reading and detecting device for detecting and recognizing a software module, receiving the The image captured by the photographic lens is used to classify and identify the traffic signal and the image of the chronograph; the mobile computing processing device processes and calculates the image classified and recognized by the software module via the traffic light and the chronograph; a voice playing software module and a voice output interface, receiving a processing result of the mobile computing processing device, selecting a corresponding voice file to be broadcasted; and wherein the traffic light and the reading device detect and identify the software module, When a countdown display is detected, a countdown recognition synchronization broadcast function is provided to synchronize the number of seconds of the countdown display with the number of seconds of the voice broadcast. A voice broadcaster for traffic signal and countdown image detection and identification, comprising: a photographic lens for capturing an image in front of driving; a traffic light and a reading and detecting device for detecting and recognizing a software module, receiving the The image captured by the photographic lens is used to classify and identify the traffic signal and the image of the chronograph; the mobile computing processing device processes and calculates the image classified and recognized by the software module via the traffic light and the chronograph; a voice playing software module and a voice output interface, receiving a processing result of the mobile computing processing device, selecting a corresponding voice file to be broadcasted; and wherein the traffic light and the countdown detection and recognition software module can be set The number of seconds to start counting down, and timely remind the driver with voice synchronization. A voice broadcaster for detecting and recognizing a traffic light number and a countdown image according to any one of claims 1 to 2, wherein the mobile computing processing device comprises a high-speed high-performance multi-core processor combined with a synergy An arithmetic processor, a controller, and a memory are connected, and an input/output interface can be connected to a photographic lens or a microphone. The voice broadcaster for detecting and recognizing the traffic light and the image of the countdown device according to any one of the claims 1 or 2, further comprising: a WiFi communication port, a Bluetooth communication port, a 3G communication port, and a 4G communication port. , as well as GPS receivers and three-dimensional gravity acceleration detectors. The voice broadcaster for the traffic signal number and the image recognition and recognition of the image recorder according to any one of the claims 1 or 2, further comprising a microphone and a voice recognition software module for voice recognition. The voice broadcaster for detecting and recognizing the traffic light and the image of the countdown device according to any one of the claims 1 or 2, further comprising a 3D image sensing chip and a back lens and a gesture detection control The software module provides a non-touch gesture control function to directly activate, select or disable the related detection and recognition function with a gesture command. A voice broadcaster for detecting and recognizing a traffic light number and a countdown image as described in any one of claims 1 or 2, wherein the traffic light and the countdown detection and identification software module are based on The road traffic sign markings issued by the government set the rules of the lights and their countdown display styles and lamp colors for detection and identification. The voice broadcaster for detecting and recognizing the traffic light and the image of the countdown device according to any one of the claims 1 or 2, wherein the voice output interface is a general wired audio jack output and a wireless Bluetooth audio output. .