This work presents a real time system of acquisition, monitoring and signal processing based on the synergy of hardware and software. In the stage of signal acquisition, a circuit board was built. This circuit board acquires two signals, one from a phototransistor sensor and other one from an ultrasonic sensor. The sending and processing of the data was through the Arduino platform. A graphical user interface (GUI) using Matlab® was implemented for the stages of monitoring and processing. For these stages two interfaces were designed and implemented with the aim of show the signals of two forms, continuous and digital, and their processing. For the processing stage, several operations such as convolution, median filter and the Fast Fourier Transform were applied to the signal. On the other hand, another aim of the implementation of this system was with the purpose to obtain a didactic tool to complement and support learning in the area of signal processing and thus generate a bigger interest and motivation of the students to practice the knowledge acquired in classes.
30 June 2021 in ECORFAN Journal-Democratic Republic of Congo
The increase in the increase in wildfires throughout the world is largely due to increases in temperature and even to an increase in the carelessness of the population in leaving a large amount of the garbage in forests. Using Python and Matlab programs were as working medium. We performed the preprocessing on multispectral images obtained by the Landsat 8 satellite with and without wildfires, which consists of three steps: alignment, characterization and normalization, with the intention of standardization the images. From obtaining the spectral signatures of wildfires and metallic structures, boxes and whiskers diagrams, Shannon entropy and mutual information from the images, there are similar behavior in bands 6, 7, 8, 10 and 11, with more relevant information, taking into account that each image is formed by 11 bands, and in bands 1, 2, 3, 4, 5, 8 and 9 there is less information, SVD decomposition allows to have the best k-rank approximation to the original data matrix. The purpose of this analysis is to reduce the computational complexity.