Implementation of Sensors Using Matlab and C Softwares

Transcription

1 Implementation of Sensors Using Matlab and C Softwares Ch.SaiSubhash M.TECH CSP Department. VR Siddhartha Engineering College, Kanuru, Vijayawada , India. S. Chitti Babu Associate Professor, ECE Dept. VR Siddhartha Engineering College, Kanuru, Vijayawada , India. ABSTRACT Satellite requires absolute attitude sensor to determine its orientation with respect to some known reference source like the Sun, the Earth, the Star, etc. There are different types of absolute attitude sensor like Sun Sensor, Earth Sensorand Star Sensor etc., which have different applications for different satellites. The Sun Sensor is one of the prime sensors for a satellite, which provides the spacecraft orientation with respect to Sun. There are different types of sun sensors available according to their application and accuracy. Sun Sensors calculates the centroids and angles of the given input data. Complete process is done by using MATLAB and C software. Same code is converted in VHDL language which can be implemented in FPGA. KEYWORDS Sun Sensor, space craft, FPGA 1.INTRODUCTION The digital sun sensor calculates the incident sunlight angle using the sunlight image registered on a CMOS image sensor. In order to accomplish this, an exact center of the sunlight image has to be determined. Therefore, an accurate estimate of the centroid is the most important factor in digital sun sensor development. The most general method for determining the centroid is the thresholding method, and this method is also the simplest and easy to implement. Another centering algorithm often used is the image filtering method that utilizes image processing. The sun sensor accuracy using these methods, however, is quite susceptible to noise in the detected sunlight intensity. This is especially true in the thresholding method where the accuracy changes according to the threshold level. In this paper, a template method that uses the sunlight image model to determine the centroid of the sunlight image is suggested, and the performance has been compared and analyzed. The template method suggested, unlike the thresholding and image filtering method, has comparatively higher accuracy. In addition, it has the advantage of having consistent level of accuracy regardless of the noise level, which results in a higher reliability[1]. One of the most important components of attitude determination system for small satellite is sun sensor. In this paper design, development and calibration of 2-axis digital sun sensor using linear Photodiodes array is presented. Senor design comprises two linear arrays mounted orthogonally to provide Sun aspect angle in two axes. The Sun aspect angle is determined by the centroid based technique and implemented using FPGA. The proposed centroid algorithm is capable of obtaining the position of the centroid to an accuracy of sub pixel level which is desirable in such designs. A method of calibration using natural sun light is also presented. Commercial grade components were used which have resulted in development of low cost, light weight and power efficient sensor[2]. 2. PROPOSED METHOD The flowchart of the proposed method is shown in Figure 1. Calculation of centroids and angles is given below. 2.1 Calculating Centroids of an Image in Matlab In order to calculate the centroid the following steps are to be followed Step1: Elimination background image from fore ground: Initially the imread reads a gray scale or color image from the file specified by the string filename. If the file is not in the current folder, or in a folder on the MATLAB path, we can specify the full pathname of that file. Global thresholding is used to eliminate the pixels with less intensity. Such that, background image with less intensity values compared to foreground of an image will be eliminated. 207

2 Mean calculates the average or mean value of an array of a foreground image. The obtained foreground image is subtracted from background image and then the resultant image is the subtracted image which contains the foreground portion of an image. 208 Figure.1: Flow chart of Image Clustering & Centroid Module Step2: Image segmentation process: Image segmentation is the process of partitioning an image into parts or regions. This division into parts is often based on the characteristics of the pixels in the image. Partitioning can be done by detecting edges of an image.

3 209 In this process, we selected Sobel edge detection which finds the edges in an input image by approximating the gradient magnitude of the image. Dilation and erosion applied on the resultant image removes the noise in the image. Dilation and erosion are done using structuring elements. Structuring element may be of different shapes. Step3: Centroid Identification: Centroid calculation of an image can be done using Region props. Region props measures centroids for each connected component in the binary image. The image is a logical array; it can have any dimension. 2.2 Centroids and Angle Calculation in C Reading data from text file Initially, text file consists the data of pixel data, pixel number and pixel intensity. The text file is opened by using fopen function. After the opening of the file, the values in the file are read out serially by using sscanf function. Figure shows the data of an image which consists of 4000 pixels Comparison of pixel intensities with threshold value While comparing pixel intensities with the threshold value, invalid pixels will be removed such that valid pixels are obtained. Initially, threshold value will be set to one value. Pixel intensities of an image are compared with the threshold value. If the pixel data is greater than the threshold value, then that pixel is considered as valid pixel and pixels with less intensity are considered as invalid pixels Identifying image clusters: All the valid pixels are considered for clustering process and the clustering process is carried out by K-means algorithm. Clustering is nothing but grouping of similar data. Kmeans algorithm depends upon the euclidean distance between two points. Clustering of data is a method by which large sets of data are grouped into clusters of smaller sets of similar data. Mean distance is calculated by the difference between the present pixel number and previous pixel number. Initially previous pixel number is initialized to zero. Distance is considered as some value. If mean distance is greater than the distance, then that pixel is added to new cluster,such that k clusters are formed. Top three clusters will be selected from k clusters. Selecting process depends upon the cluster with more pixel size Centroid Calculation For the three clusters centroids are calculated. The three centroids are calculated by using momentum method. Centroid = pixel number * pixel intensity / Pixel intensity The sum of the product of the pixel number and the pixel intensity for the pixel (i.e. S I * X) is accumulated in a variable pixel_cnt [count] which is an array of size 4000 where the index of the array is the cluster number. Where I is the intensity of the current pixel and X is the co-ordinate of the current pixel. Similarly the sum of the intensities of the pixels (S I) are stored in the variable pixel_data [count] which is an array of size 4000 where the index of the array is the cluster number.the three centroids are calculated by momentum method by dividing the centroid numerator by the centroid denominator for the same cluster_number and stored in variable centroid[i]. Centroid[i] = pixel_cnt[i] /pixel_data[i] After getting the three centroids, they are arranged according to co-ordinate numbers and then Angles are then calculated Angle calculation This three centroids information is converted to two sun aspect angles using transfer function. This transfer function consists of trigonometric equation (Tan function) which can implemented in code blocks. First we have to calculate the central spot of centroid. Initialize with three different values. delta_y1 = image_selc [1]-1000; delta_y2 = image_selc [2]-2000; delta_y3 = image_selc [3]-3000; And then calculate the central spot dev_y0 and dev_y1: dev_y0 = delta_y2; dev_y1 = (delta_y1+delta_y3)/2;

4 After calculating central spots between images, calculate sun angle using trigonometric Function. The transfer equation is given below: x_angle = ((180*7)/22)*atan (dev_y0/3); y_angle = ((180*7)/22)*atan ((dev_y1-dev_y0)/3); Where dev_y0 and dev_y1 are the central spots, And the height of detector is taken as 3, And x_angle&y_angle are the two sun angles of the data. 3.SIMULATION RESULTS The simulation results using matlab and C softwares are shown below Simulation Results Using Matlab Simulation results using matlab is shown below. Figure 2 represents input image. Figure 3 represents foreground image and segmented images. Figure 4 represents centoid of an image. Figure.2: Input Image Figure.3: Foreground Image and Segmented Images 210

5 Figure.4: Centroids of an Image 3.2. Simulation Results Using C Simulation results using C software is shown below. Figure 5 represents opening and reading a text file. Figure 6 represents pixel intensities greater than the threshold value. Figure7 represents identifying image clusters. Figure8 represents centroids and angle calculation. Figure.5:Opening and Reading a Text File 211

6 Figure.6: Pixel Intensities Greater than the Threshold Value Figure.7:Identifying Image Cluster 212

7 Figure.8: Centroids and Angle Calculation 4. CONCLUSION: In this paper, we present the results of the Sun Sensor algorithm. Micro Sun Sensor is a novel two axes high accuracy, compact sun sensor to meet the Sun sensor requirements of micro/nano satellites. The complete algorithm is developed and simulated in MATLAB and C language. REFERENCES [1] Young-Keun Chang, Byung-Hoon Lee and Seok-Jin Kang, High-Accuracy Image Centroiding Algorithm for CMOS-Based Digital Sun Sensors IEEE sensors , [2] Bilal A. Alvi, Naeem Abbas, Amber Israr and Muhammad Asif, Optimized Design of Sun Sensor and Centroid Algorithm for Small Satellite Mission conference-wireless VITAE [3] AmitMaji, Veena John: High speed real time image clustering and centroid computing for active pixel sensor based sun sensor XXXV OSI, symposium