| The fire helmet platform is designed in this article based on image enhancement technology,which is mainly controlled by the i TOP-4412 processor,and uses the Heyman thermal imaging sensor HTPA120x84 d to collect flames that have not been extinguished in high-temperature areas.It can enable firefighters to see flames in high-temperature areas in the presence of smoke and ensure their personal safety.Due to the low resolution and low contrast of grayscale images is collected by thermal imaging sensors,firefighters find it difficult to distinguish flames in high-temperature areas from complex backgrounds.Therefore,C++/C language+Open Cv is used on the embedded platform’s Qt Creator to enhance the grayscale images.Firstly,compared with traditional Canny object detection algorithms,a Canny detection algorithm that integrates the improved OTSU alorithm is proposed to extract flames,resulting in flame contours with more continuous edges and more edge information.In order to present the flame contour more intuitively and clearly to firefighters,a frequency domain pseudo color enhancement based on HSI space is used to enhance the flame contour.The obtained flame contour is more intuitive,with more detailed information and more natural colors.Finally,the image of the fire scene is obtained by the RGB camera fused with the pseudo color enhanced flame contour at the pixel level,and the fused video image is sent to the remote monitoring room through Wi Fi network based on TCP protocol,and displayed on the LCD of the development board.This article adopts a relatively novel design approach.The bottom layer of the entire platform design is based on ARM to achieve image acquisition and hardware drive.The upper layer is applied in embedded systems through research on image enhancement algorithms and Wi Fi network based video wireless transmission.The platform is modularized using multithreading technology,and the functions of each module are ultimately implemented. |
PDF Full Text Request