1. Field of the Invention
The present invention relates to traffic sign identification technology and more particularly, to a traffic sign identification system and method that uses a color space conversion technique and image convergence conversion to obtain a skeleton diagram for the identification of a traffic sign.
2. Description of the Related Art
With the progress of the society, cars and buses have become the main transport vehicle of most people. In order to accurately guide the driver to the destination, car navigation technology has been created. Conventional car navigation techniques are commonly of a satellite-based navigation system that uses a satellite receiver to receive signals from multiple satellites for allocating the current location of the car through an algorithm so that related landscape information, speed restriction and other traffic information can be provided to the driver, effectively solving the path-finding problem when the driver is driving the car.
However, conventional car navigation systems can simply guide the driver to drive the car only in accordance with the path planned, and remind the driver of related traffic signs with the information provided in the database. When encountered a speed limit change in an unfamiliar road, the drivers are often unable to respond immediately, resulting in speeding. So, how to be able to pay attention to traffic signs on the route under limitations from the environment and to get complete and effective navigation information is a serious problem to be solved.
The present invention has been accomplished under the circumstances in view. It is the main object of the present invention to provide a traffic sign identification system and method, which employs an image recognition technique to obtain any traffic restrictions available at any time on the road, effectively advising the driver to avoid speeding.
To achieve this and other objects of the present invention, a traffic sign identification system in accordance with the present invention comprises a traffic sign image capture unit adapted to capture a road condition image, a traffic sign image recognition unit adapted to pick up a traffic sign image from the road condition image captured by the traffic sign, an arithmetic processing unit adapted to convert the original RGB color space of the traffic sign image into a HIS (Hue-Intensity-Saturation) color space, to extract one or multiple graphical features from the HIS color space subject to a predetermined graphic pattern, to execute a positioning procedure to allocate one traffic sign in the traffic sign image so as to further obtain a digital speed limit image using a matching arithmetic, and to convergence-convert the digital speed limit image into a digital speed limit skeleton diagram for identification of the speed limit value of the traffic sign image.
Preferably, the traffic sign image capture unit is selected from the group of cameras, video cameras, driving recorders and mobile electronic devices having a camera function.
Preferably, the arithmetic processing unit comprises a color conversion module, a feature extraction module, and a skeleton conversion module. The color conversion module is adapted to convert the original RGB color space of the traffic sign image into a HIS (Hue-Intensity-Saturation) color space. The feature extraction module is adapted to extract one or multiple graphical features from the HIS color space subject to a predetermined graphic pattern, and to execute a positioning procedure to allocate one traffic sign in the traffic sign image so as to further obtain a digital speed limit image using a matching arithmetic. The skeleton conversion module is adapted to convergence-convert the digital speed limit image into a digital speed limit skeleton diagram for identification of the speed limit value of the traffic sign image.
To achieve this and other objects of the present invention, a traffic sign identification method in accordance with the present invention comprises the step of: operating a traffic sign image capture unit to capture a road condition image, the step of operating a traffic sign image recognition unit to pick up a traffic sign image from the road condition image captured by the traffic sign, the step of operating an arithmetic processing unit to convert the original RGB color space of the traffic sign image into a HIS (Hue-Intensity-Saturation) color space, the step of extracting one or multiple graphical features from the HIS color space subject to a predetermined graphic pattern and executing a positioning procedure to allocate one traffic sign in the traffic sign image so as to further obtain a digital speed limit image using a matching arithmetic, and the step of convergence-converting the digital speed limit image into a digital speed limit skeleton diagram for identification of the speed limit value of the traffic sign image.
Preferably, the step of operating an arithmetic processing unit to convert the original RGB color space of the traffic sign image into a HIS (Hue-Intensity-Saturation) color space is to operate a color conversion module of the arithmetic processing unit to convert the original RGB color space of the traffic sign image into a HIS (Hue-Intensity-Saturation) color space.
Preferably, the step of extracting one or multiple graphical features from the HIS color space subject to a predetermined graphic pattern and executing a positioning procedure to allocate one traffic sign in the traffic sign image so as to further obtain a digital speed limit image using a matching arithmetic is to operate a feature extraction module for extracting one or multiple graphical features from the HIS color space subject to the predetermined graphic pattern and then to execute a positioning procedure to allocate one traffic sign in the traffic sign image so as to further obtain a digital speed limit image using a matching arithmetic.
Preferably, the step of convergence-converting the digital speed limit image into a digital speed limit skeleton diagram for identification of the speed limit value of the traffic sign image is to operate a skeleton conversion module for convergence-converting the digital speed limit image into a digital speed limit skeleton diagram for identification of the speed limit value of the traffic sign image.
In order to increase the recognition rate without increasing the effectiveness burden, the invention uses a traffic sign image capture unit to capture a part of the road condition image, utilizes the features in the image to determine the speed limit sign to be on the left or right side of the road, and tactically moves the identification range. Thus, the invention can effectively maximize the detection area and improve the recognition rate without changing the range of identification.
Other advantages and features of the present invention will be fully understood by reference to the following specification in conjunction with the accompanying drawings, in which like reference signs denote like components of structure.
Referring to
The arithmetic processing unit 130 comprises a color conversion module 140 adapted for efficient implementation of color space conversion from RGB color space to HIS (Hue-Intensity-Saturation) color space, a feature extraction module 150 adapted to extract one or multiple graphical features from the HIS color space subject to a predetermined graphic pattern, to execute a positioning procedure to allocate one traffic sign in the traffic sign image so as to further obtain a digital speed limit image using a matching arithmetic, and a skeleton conversion module 160 adapted to convergence-convert the digital speed limit image into a digital speed limit skeleton diagram for identification of the speed limit value of the traffic sign image.
In this embodiment, the traffic sign image capture unit can be a camera, video camera, driving recorder, or a mobile electronic device with a camera function.
Referring to
S210: Operate a traffic sign image capture unit to obtain a road condition image.
S220: Operate a traffic sign image recognition unit to pick up at least one traffic sign image from the road condition image.
S230: Operate a color conversion module of an arithmetic processing unit to convert the original RGB color space of each traffic sign image into a HIS (Hue-Intensity-Saturation) color space, as shown in
S240: Operate a feature extraction module to extract one or multiple specific graphical features from the HIS (Hue-Intensity-Saturation) color space subject to a predetermined graphic pattern, and to execute a positioning procedure to allocate one traffic sign in the traffic sign image so as to further obtain a digital speed limit image using a matching arithmetic, as shown in
S250: Operate a skeleton conversion module to convergence-convert the digital speed limit image into a digital speed limit skeleton diagram for identification of the speed limit value of the traffic sign image, as shown in
The aforesaid skeletonization is to reduce the storage space of the space. All graphics have become a pixel (the width of 1-pixel). Therefore, any intersection of lines can be determined to be a pixel, and then skeleton conversion processing is performed so as to obtain a digital speed limit skeleton 300D.
In conclusion, the use of a circular template of block coding and the introduction of statistical conception in accordance with the present invention enable any circular object in a poor quality image to be matched and allocated. The invention runs a skeletonization algorithm on the content of the traffic sign and applied an object pixel ratio method for the calculation of a particular block, successfully identifying digital computation at a lower volume and improving traffic sign recognition convenience.
Although a particular embodiment of the present invention has been described in detail for purposes of illustration, various modifications and enhancements may be made without departing from the spirit and scope of the invention. Accordingly, the invention is not to be limited except as by the appended claims.