This section describes the algorithm used for traffic light and road sign detection. It also provides the hardware components used to provide output. In the proposed system, initially input image was obtained in RGB form and converted to gray scale and finally to binary images. This said to be pre-processing of images. Then filtering was performed for removal of noise from images. Color of traffic light was detected based on threshold values. Hough transform was done for the detection of shape and based on the medval the road signs are classified. Fig 1 represents the work flow of the proposed system.
3.1 Pre-processing of images
Before extracting information from the image, it is necessary to apply some pre-processing procedures. The first step is to convert RGB image to gray scale image. Gray scale transform is most widely used if the processed output is viewed by human. The gray scale image carries only the intensity information and it have only black and white with gray shadows in between. The gray scale weighted average can be given by, X=0.299r+0.587g+0.114b (1) Here r,g,b represents the values of red, green and blue components. 3.2 Gaussian filtering Gaussian filtering acts like a low pass filter that suppress high frequency details like noise and preserves low frequency parts of image. This low frequency parts have the required information and because of the noise removal it would be easy for identification of specified region in the image. Fig 1 Flow chart for proposed system The Gaussian filter function is given by, G(x, y) = (2) The value of sigma represents the inverse of amount of filtering. Smaller value of sigma indicates larger amount of frequencies are supressed. 3.3 Color analysis Based on threshold values the color of the traffic light glowing has been detected. Generally there are three color components: red, blue and green. The count values rcount, bcount and gcount are obtained and based on this values the color has been identified. 3.4 Hough transform Hough transform is a feature extraction technique in which the shape of road sign is identified. The classification of the sign is done on the basis of the shape of road sign. Initially the median filtering has …show more content…
Fig 2 Block Diagram of proposed system
3.7 Arduino UNO
Arduino is an open source platform mainly used for electronic projects. Arduino can easily interact with LEDs, motors, camera, speakers and even with the smartphones. So it is a most preferred development platform for all the applications. The arduino provides all the controller of ATmega from ATMEL company.
Arduino UNO is great and most preferred choice for the development. It consists of 14 digital I/O pins, 6 analog inputs, a usb connection, a power jack and a reset button. The Arduino IDE provides a software library, which provides procedure for common input and output. Arduino also provides support for both C/C++ languages.
3.8 Voice module
Arp R33 voice board is provided for the voice output. Arp provides two modules: dual oscilloscope and voltage controlled voice module. The main VCO board of Arp R33 provides a 1.5kohm resistor. It provides eight channels and can support up to eight voice output. The voice can be recorded up to one minute and it can be easily interfaced with Arduino. Hence, the voice module can easily implemented and output can be
3.1 Pre-processing of images
Before extracting information from the image, it is necessary to apply some pre-processing procedures. The first step is to convert RGB image to gray scale image. Gray scale transform is most widely used if the processed output is viewed by human. The gray scale image carries only the intensity information and it have only black and white with gray shadows in between. The gray scale weighted average can be given by, X=0.299r+0.587g+0.114b (1) Here r,g,b represents the values of red, green and blue components. 3.2 Gaussian filtering Gaussian filtering acts like a low pass filter that suppress high frequency details like noise and preserves low frequency parts of image. This low frequency parts have the required information and because of the noise removal it would be easy for identification of specified region in the image. Fig 1 Flow chart for proposed system The Gaussian filter function is given by, G(x, y) = (2) The value of sigma represents the inverse of amount of filtering. Smaller value of sigma indicates larger amount of frequencies are supressed. 3.3 Color analysis Based on threshold values the color of the traffic light glowing has been detected. Generally there are three color components: red, blue and green. The count values rcount, bcount and gcount are obtained and based on this values the color has been identified. 3.4 Hough transform Hough transform is a feature extraction technique in which the shape of road sign is identified. The classification of the sign is done on the basis of the shape of road sign. Initially the median filtering has …show more content…
Fig 2 Block Diagram of proposed system
3.7 Arduino UNO
Arduino is an open source platform mainly used for electronic projects. Arduino can easily interact with LEDs, motors, camera, speakers and even with the smartphones. So it is a most preferred development platform for all the applications. The arduino provides all the controller of ATmega from ATMEL company.
Arduino UNO is great and most preferred choice for the development. It consists of 14 digital I/O pins, 6 analog inputs, a usb connection, a power jack and a reset button. The Arduino IDE provides a software library, which provides procedure for common input and output. Arduino also provides support for both C/C++ languages.
3.8 Voice module
Arp R33 voice board is provided for the voice output. Arp provides two modules: dual oscilloscope and voltage controlled voice module. The main VCO board of Arp R33 provides a 1.5kohm resistor. It provides eight channels and can support up to eight voice output. The voice can be recorded up to one minute and it can be easily interfaced with Arduino. Hence, the voice module can easily implemented and output can be