The usage of adaptive threshold method in the system of computer vision

The article touches upon the question of the edge detection of digital images in robotics and other fields. The algorithm of the edge detection called ‘Adaptive Threshold’ serves as the subject matter of the research. This algorithm has been described in details and then, tested on different types of images. The results of the research done have shown that this method is efficient enough for images with small amount of noise reduction. However, the quantity of fidelity in a resulting image decreases with the increasing amount of noise. The study has concluded that before processing it is preferably to employ the median filtering to reduce the noise level and thus, to improve the algorithm output. It has been found that the algorithm is rather universal and is suitable for processing a wide range of images.