Research On Digital Water Level Sensor Based On Image Processing

With the continuous development of modern information technology, modern information technology is widely applied to all walks of life in the society.Water conservancy industry safety, is an important factor for the prosperity and development of the society,as well as people’s stable life. The application of modern information technology can better ensure the safety of water resources utilization of water resource, prevent and control the frequent occur of flood disasters.In this paper, water level detection system was designed based on ARM microprocessor and visual sensors. There are advantages of non-contact detection system, as well as a wide range of applications, and the characteristics of easy installation and unattended, in traditional vision sensors, the detected images were obtained by the image input device,and then would be sent to the PC for processing. In this article, firstly, the image signal collected would be collected by the vision sensor,and then would be processed and detected in the microprocessor,as avoid the interference in the process of transmission. Secondly, image quality would be improved by image enhancement technology to reduce the influence of the various types of noise in the image, and then the feature amount might be more apparent through the gradation equalization and image binarization. After that, extract characteristic information by edge detection. In this article, different water image processing results by Robert operator, sobel operator, prewitt operator and the Laplace operator were discussed.combined with real-time requirements of the system, robert operator,which is suitable for processing effects and processing speed, was chosen for edge detection. At last, extract the water line and the calibration line of the image by straight line extraction of hough transform. Finally,the actual distance can be converted by the numerical value of the distance,which was calculated through Machine vision theory and method,as well as the projection relation between the CCD camera and the objective world.At the same time, this paper presents the hardware design ideas. Firstly,the main hardware selection for system and the design of the associated circuit were all analyzed in this article,including a microprocessor, reset circuit, clock circuit, and the A/D serial. Secondly, the dedicated solar power system and low power consumption were designed,which manage power to the external electrical devices,which don’t need to always be available,through that ARM pin control transistor conduction, for the system characteristic which is special for the field unattended work. Thirdly, because of the way for field work in this paper, so the selection of the transmission mode is unlimited network transmission. Meanwhile, the data transmission in the system is digital signal, so in the choice of the transmission method,the author did not choose the high-speed 3G network,but the GPRS network. Although the speed of the GPRS network is lower than the ones of the 3G network,the coverage rate of the GPRS network is greater than the coverage of the 3G network, which can meet the demand of system transmission.In addition, the working principle and hardware structure of the CCD camera were studied in the article. The key factors,the suitable light sources and lighting, decide whether the CCD camera can get a high-quality image. Finally, through discussing in detail the light source and lighting used by the CCD camera,a suitable light source and lighting were chosen for the present system.Finally, the simulation and the real-time nature of the present system was tested, the desired results were obtained.