Design And Implementation Of The Embedded System Of An Endoscope-type Video Search And Rescue Unit

When natural or man-made disaster occur causing the collapse of large constructions, people may remain trapped under the ruins. To search and rescue victims in time requires obtaining their exact location, identifying the environment around them and assessing their physical condition. Most such problems could be eased with accurate video information of the people concerned.Inspired by industrial endoscope, this paper designed an endoscope-type video search and rescue unit composed of a scalable pole with a control box at the bottom. A cradle head consisting of two motors is installed at one end of the pole. The two motors rotate in different axial directions. A CMOS camera is installed inside the cradle head, driven by the two motors, that covers all view angles. An LCD touch screen, an ARM11control board and a battery are set inside the control box. Video information collected by the camera can be shown directly on the screen. By touching the screen, the video can be encoded into H.264format files and saved in the SD card through a multi-format-codec module inside the ARM11. With this module, the saved video can also be decoded and shown on the screen. Besides, it can save the video data into JPEG format, and display the pictures conveniently. By pressing the button in the control box, the user can control the rotation of the motors.The work for this paper mainly consisted of two parts:designing the Video&Motor control software, and designing the hardware circuit of the control box.For the hardware part, we chose the UT-6410core board; it is based on a Samsung S3C6410type ARM11chip. The design includes power circuit, motor control circuit, camera interface circuit, LCD interface circuit, SD card interface circuit, serial port circuit, the design and debugging of the PCB board. The software part, based on embedded Linux operating system, has functions of video real-time displaying, video encoding and saving, video decoding, picture encoding and picture decoding. Written with Qt, the software has a friendly user interface. To control the two motors, we designed the appropriate driver and the control program in the upper layer.