— The use of digital instruments in industries and laboratories is rapidly increasing as they are easy to calibrate and have relatively high accuracy. In this paper, an automatic data acquisition system using the OCR technique from digital multimeters and other similar digital display devices is proposed. The input image is taken from a digital multimeter with a seven-segment LCD display using a web cam. The image is then processed to extract numerical digits which are recognized using a feedforward neural network. The recognized values ​​can then be exported to a spreadsheet for graphing and further analysis. A distinct advantage of this method is that it can automatically detect the decimal point and negative sign. This setup can be used in real-time systems employing a wide variety of digital display tools, with high accuracy. Keywords: OCR, adaptive thresholding, data acquisitionI. INTRODUCTIONMeasuring devices in the laboratory and in some industries are usually equipped with a display unit through which the result is detected. The process of collecting data output from the display is usually done manually or via data acquisition cards which may not always be available. Furthermore, commonly available data acquisition cards are very expensive. We therefore propose a solution to this data acquisition problem using a webcam and a processing unit that can be provided at a nominal cost. It also can eliminate the error of the human eye and adjust the reading range. The main goal of the OCR technique is to distinguish the object of interest from the original image. To achieve this, the noise within the image must be removed up to a certain level without distorting the character located in the center of the paper...... SJ Perantonis, B. Gatos and N Papamarkos "Image segmentation and identification of linear features using rectangular block decomposition” ICECS '96 Proceedings of the Third International Conference on Electronics, Circuits and Systems, pp-183-186.[13] Shaoyuan Sun, Haitao Zhao “Kernel Averaging Filter” 2008. CISP '08 Conference on Image and Signal Processing, pp-681-685.[14] Derong Liu, Tsu-Shuan Chang, and Yi Zhang, "A constructive algorithm for incrementally trained feedforward neural networks", IEEE Transactions on Circuits and Systems I: Fundamental Theory and Applications, vol - 49 no - 12, pp - 1876-1879. [15] Nallasamy Mani and Bala Srinivasan. “Application of Artificial Neural Network Model for Optical Character Recognition” 1997 IEEE International Conference on Systems, Man and Cybernetics and Simulation, vol - 3, pp-2517-2520.