Many deep-learning computer vision systems analyse objects not previously observed by the system. However, such tasks can be simplified if the objects are marked beforehand. A straightforward method for marking is printing 2D symbols and attaching them to the objects. Selecting these symbols can affect the performance of the CV system, as similar symbols may require extended training time and a larger training dataset. It is possible to find good symbols differentiated by the given neural network easily. Still, there were no efforts to generalise such findings in the literature, and it is not known if the symbols optimal for one network would work just as well in the other. We explored how transferable symbol selection is between the networks. To this end, 30 sets of randomly selected and augmented symbols were classified by-five neural networks. Each network was given the same training dataset and the same training time. Results were ranked and compared, which allowed the identification of networks which performed similarly so that the symbol selection could be generalised between them.

Deep learning techniques in computer vision (CV) tasks such as object detection, classification, and tracking can be facilitated using predefined markers on those objects. Selecting markers is an objective that can potentially affect the performance of the algorithms used for tracking as the algorithm might swap similar markers more frequently and, therefore, require more training data and training time. Still, the issue of marker selection has not been explored in the literature and seems to be glossed over throughout the process of designing CV solutions. This research considered the effects of symbol selection for 2D-printed markers on the neural network’s performance. The study assessed over 250 ALT code symbols readily available on most consumer PCs and provided a go-to selection for effectively tracking n-objects. To this end, a neural network was trained to classify all the symbols and their augmentations, after which the confusion matrix was analysed to extract the symbols that the network distinguished the most. The results showed that selecting symbols in this way performed better than the random selection and the selection of common symbols. Furthermore, the methodology presented in this paper can easily be applied to a different set of symbols and different neural network architectures.

—The problem of transport optimization is of great importance for the successful operation of distribution companies. To successfully find routes, it is necessary to provide accurate input data on orders, customer location, vehicle fleet, depots, and delivery restrictions. Most of the input data can be provided through the order creation process or the use of various online services. One of the most important inputs is an estimate of the unloading time of the goods for each customer. The number of customers that the vehicle serves during the day directly depends on the time of unloading. This estimate depends on the number of items, weight and volume of orders, but also on the specifics of customers, such as the proximity of parking or crowds at the unloading location. Customers repeat over time, and unloading time can be calculated from GPS data history. The paper describes the innovative application of machine learning techniques and delivery history obtained through a GPS vehicle tracking system for a more accurate estimate of unloading time. The application of techniques gave quality results and significantly improved the accuracy of unloading time data by 83.27% compared to previously used methods. The proposed method has been implemented for some of the largest distribution companies in Bosnia and Herzegovina.