Search

The subject of this study is the pixel-by-pixel controlled classification of multichannel satellite images distorted by additive white Gaussian noise. The paper aim is to study the effectiveness of various methods of image classification in a wide range of signal-to-noise ratios; an F-measure is used as a criterion for recognition efficiency. It is a harmonic mean of accuracy and completeness: accuracy shows how much of the objects identified by the classifier as positive are positive; completeness shows how much of the positive objects were allocated by the classifier. Tasks: generate random valuesof the brightness of the noise components, ensuring their compliance with the accepted probabilistic model; implement the procedures of element-wise controlled classification according to the methods of support vectors, logistic regression, neural network based on a multilayer perceptron for images distorted by noise; evaluate and analyze the results of objects bezel-wise classification of noisy images; investigate the effect of noise variance on classification performance. The following results are obtained. Algorithms of pixel-by-pixel controlled classification are implemented. A comparative analysis of classification efficiency in noisy images is performed. Conclusions are drawn. It is shown that all classifiers provide the best results for classes that mainly correspond to areal objects (Water, Grass) while heterogeneous objects (Urban and, especially, Bushes) are recognized in the worst way; classifiers based on the support vector machine and logistic regression show low recognition accuracy of extended objects, such as a narrow river (that belongs to the wide class of "water"). The presence of noise in the image leads to a significant increase in the number of recognition errors, which mainly appear as isolated points on the selected segments, that is, incorrectly classified pixels. In this case, the best value of the classification quality indicator is achieved using neural networks based on a multilayer perceptron.

Existing methods of quality assessment of flat products are not sufficiently effective: they can be used for technological process management with certain limitations. Our research suggests an improved approach to the calculations of priority number of risks. The new suggested approach will allow risk assessment of dependent undesired events.

The article discusses the models of repair process of a turbine compressor package blades. The author determines the boundaries of the assured quality area considering mathematical model analysis and defect detection of rotor blades based on the results of natural vibration frequency measurements.

Modern databases are crucial components of quality management systems according to the requirements of ISO 9001:2015, but their design and maintenance need significant resources. Despite the fact that modern databases are a key factor of management decision making, not all organizations can spend sufficient resources on their development what results in the necessity to find a model for database maintenance in the conditions of limited resources.

Human factor plays a crucial role at a company. Thus, it is important to consider and forecast its reaction (such as resistance) at implementation of changes. Our paper suggests an approach which compares models of new management standards to be introduced with key concepts of a management system as seen by management staff. If we define the coordination degree between management model suggested by an official standard and conceptual subjective model of a member of managerial staff, we can make a conclusion about the resistance degree to changes. Assessment of the resistance degree and evaluation of resistance causes allows us organizing special training aimed at elimination of unharmonious and conflictive aspects of new standards. As a result of the training the most compromise pilot project can be chosen to valuate changes.

Small unmanned aerial vehicles (SUAVs) have recently attracted considerable research and development interest owing to their small size, portability, and limited operational risk. In SUAVs, the Reynolds number Re is typically very low because they are small and fly at low speeds. The Reynolds number strongly influences the airfoil performance, and at Re = 105, the lift-drag (L/D) ratio decreases significantly and the airfoil performance deteriorates. Since SUAVs must be able to carry payloads such as communication devices and observation equipment, the challenge is to maintain sufficient payload capacity. We propose a high-lift wing that can maintain a high L/D ratio and increase more payload in the low Reynolds number region (Re = ~105) by small size airfoil (SUAV), thereby expanding the application range of SUAVs. A high L/D ratio is achieved by converting the top and bottom face of the wing into a driven belt, thereby actively adding momentum to the flow around the wing. A test was conducted at a Reynolds number of 1.1 × 105 chord. The maximum L/D ratio increased by a factor of 1.67. A flow visualization test was also conducted at a Reynolds number of 105. Flow attachment was observed up to an angle of attack of 28° with the device; in comparison, the flow separated at an angle of attack of 11° in a non-driven wing.