Research on the automated generation method of laboratory testing report

In addressing the issues of ambiguity and volatility in collecting value at a specific moment as testing data during automated laboratory testing report generation, the method is proposed by borrowing from manual operational procedures by adopting a sliding window to continuously collect data, and the ADF (Augmented Dickey-Fuller) method is utilized to automatically determine the stability of the data, and the maximum-minimum difference method should be applied to set limits on the fluctuation range, and calculate the average value as the testing data. The morphological algorithm is employed to highlight the features of text in testing picture to reduce the impact of color and texture features of the testing sample, and OCR (Optical Character Recognition) technology is adopted to identify the text and its orientation. By reconstructing the coordinate axes to determine the quadrant where the text is located, combined with prior knowledge, the correct orientation of the testing picture is judge, which overcomes problem that the characteristics of the testing picture are complex and the text orientations is not unified, which lead to the overall orientation is difficult to judge. Consequently, this is convenient to insert the testing report, improves work efficiency, and alleviates the "slow to test" issue in the testing industry.