Assisted Analysis Of Telegraph Key Transmitting Training Based On Image Processing

How to train recruits to master the skills of telegraph key transmitting is always an important and difficult task for communication troops. In order to help raw telegraphers to perceive their wrong movements and correct them just in time, and to help instructors to grasp the training situation comprehensively, objectively and accurately, we propose the idea of applying the image processing technology to the telegraph training and demonstrate the feasibility of the idea theoretically and technically.Instead of the qualitative judgment by experience, we can make a quantitative analysis of transmitting movements by datum extracted from the processed images of hand movements. For this purpose, firstly, we analyze the regular pattern of hand movements in transmitting and find the key graphic features which can be utilized to judge and evaluate the hand movements. Secondly, we use two methods to separate the area which includes the key features from the image respectively, and compare their advantage and disadvantage with each other. One method is complexional segmentation and the other is sequential images difference. Thirdly, we process the separated area to extract the key features from it. Finally, we integrate the methods and steps mentioned before into a simulation program, and extract the key features practically from a telegraph video. Using the key features, we can calculate datum to judge and evaluate the hand movements in transmitting.This paper is an attempt to bring the image processing technology into the military training field. It's also an exploration of standardizing and scientizing military training that making a quantitative analysis of some professional skills.