digital images

Methods for Recognition of Weapons of Preliminary Work up Digital Images of Traces on Bullets and Cartridge Cases

In paper investigated methods to improve the quality of digital images of the firing pin traces and traces on fired bullets. Methods of comparison of firing pin traces by means of Euler’s characteristics and methods of comparison of traces on fired bullets by means of the correlation analysis have been considered. Analyzed the firing pin marks on the image cartridges, shot in the Makarov pistol and Colt mod. 1911, as well as marks on the bullets, shot from a Makarov pistol.

Classifi cation of fi ring pin marks images by weapon specimens using a fully-connected neural network

Introduction. The aim of the work is to increase the effi ciency of identifi cation of fi rearms by images of fi ring pin marks in the automatic mode. The relevance of the task is determined by the low effi ciency of the known methods of automatic identifi cation of fi rearm by the fi ring pin marks with individual topological types of individualizing features. This aff ects the investigation of crimes related to the use of fi rearms. Formation of clone images. A training sample was formed; it included 140 original images of fi ring pin marks from 50 classes, on the basis of which about 1000 clone images were made with slightly modifi ed individualizing features. In this case a specifi c specimen of a fi rearm is meant as a class. Neural network training. A fully connected neural network with the following architecture was used as a classifi er: an input layer of neurons; two hidden layers; an output layer. The input layer included 2500 neurons, the fi rst hidden layer was made up of 625 neurons, the second hidden layer contained 156 neurons; the output layer consisted of 50 neurons (in accordance with the number of the classes). Evaluation of the calculation results. The prediction accuracy of the trained neural network was estimated according to the Accuracy metric, which is the ratio of the number of correct predictions to the total number of predictions. The prediction accuracy for the maximum signal on one output neuron was 81%, and when the maximum signals on three output neurons were taken into account, the accuracy was about 91%. Conclusions. The research has shown the possibility of classifi cation of the images of fi ring pin marks by weapons using a fully connected neural network, as well as the eff ectiveness of using artifi cially generated clone images of fi ring pin marks for training a fully connected neural network in cases with a small number of initial objects.

The Concept of Mathematical Model of the Assessment of Uniqueness of Sets of Coinciding Routes in Secondary Traces on the Shot Bullets

Introduction. The model of an assessment of probability of casual combination of sets of routes in secondary traces is considered in article. Development of quantitative criteria of justification of a categorical positive conclusion about criminalistic identity of the compared traces, and also algorithm of formation in the automatic mode of the priority list is the purpose of the conducted research. Theoretical part. Twodimensional images of traces of rifling fields used for the simulation. The formulas for estimating the probability of an accidental match of trails were obtained. Experimental part. Calculations performed on the developed formulas, the dependence of the probability of an event compared to the number of trails should be shown. Conclusion. The theoretical possibility of estimating the probability of a random alignment sets of trails (the degree of matching features unique complexes) and its use in practice is shown in this paper.

The Comparison of Digital Images of Firing Pin with a Dominant Features in the Form of Circles and Arcs

Introduction. Development of the algorithms for the automatic comparison of the digital images of firing pin is an important task of the forensic examination. This task is aimed at improving the efficiency of crime investigation involving the use of firearms. In this paper images of firing pin with the features in the form of circles and arcs with single center are investigated. The method based on modified Euclidean distance between circles of the similar firing pin was proposed to assess the degree of similarity of the traces. Preliminary processing. To eliminate the adverse effect of noise and various image artifacts were pretreated. Markers were placed to accurately determine the features in the form of circles. Method of searching paired tracks. Criteria based on modified Euclidean distance were developed for the formation of the priority list. Numerical experiment. In the numerical experiment database of 60 objects was used. Paired trace from firing pin array was included in top four of the prior list in 90 percent cases. Conclusions. The proposed algorithm allows to sort effectively and quickly the array of test objects by the degree of similarity of signs in the form of arcs and circles with relevant features of the test track.