<![CDATA[Road lines with longitudinal line types are road lines that we often encounter on highways, all highways always have this type of marker. Although this marker has the function of dividing the road into two directions and helping motorists drive in place, in fact there are still many traffic accidents caused by not concentrating and are not aware that the driver is in the wrong position. On this basis as an effort to improve driving safety in combination with current computer technology, this study was made a tool that can detect the type of longitudinal road lines with the addition of features to measure the position of the vehicle with markers and provide a warning to driving if the position is in a position danger or violating line markings. Enter from the system in the form of video obtained from the use of a camera mounted on the body of a motorcycle that is parallel to the front wheels of the vehicle. The method used to recognize objects uses a Thereshold blend with Hough Transform so that an image will be produced where longitudinal line can be detected and can be calculated in length with the help of Raspberry Pi 3, then the results of the process will be displayed on the LED lights and sounds from buzzer as output media. From the results of functional testing the tool can detect markers objects and can provide information to users about the distance of vehicles with line markers, as well as sound signals if the driver is in a dangerous condition. In accuracy testing obtained an accuracy of 85% of the video sample with four distance conditions and two type longitudinal road lines conditions. From testing computation time starting from object recognition to processing data takes an average of 1,265 seconds.]]>
Copyrights © 2019
