Dongguang Li

Ballistics firearms identification based on images of a cartridge case and projectile

A spent cartridge case exhibits characteristic markings (firearm fingerprint) that can be used to identify the type, and possibly make, of weapon in which the cartridge was fired. This paper details research into the use of discriminant analysis for the purpose of matching spent cartridge cases to specific make and model of firearms. A fired projectile can also be analyzed by measuring the class characteristics such as land mark, groove, mar, and twist ratio. A personal computer (PC) based prototype software application called FireBall will also be described in detail.

A Firearm Identification System Based on Neural Network

In this paper, a Firearm Identification system based on Self-Organizing Feature Map (SOFM) neural network is proposed. We focus on the cartridge case identification of rim-firing mechanism. Experiments show that the model proposed has high performance and robustness by integrating the SOFM neural network and the decision-making strategy. This model will also make a significant contribution towards the further processing, such as the more efficient and precise identification of cartridge cases by combination with more characteristics on cartridge cases images.

Optical thin film optimization design using genetic algorithms

Optical thin films are used in a wide variety of optical components. An important aspect of modern thin film design work is the use of computers to match the multilayer parameters to a set of optical specifications such as a desired reflectance curve. There are several basic approaches to the design of thin film multilayer coatings. These include graphical, analytical and digital design methods. The latter, representing both local and global minimum seeking algorithms, are particularly powerful because they lend themselves to the design of coatings with much more complicated properties than is possible with the other methods. Many traditional optimization techniques, including Simplex, Gradient, and Damped least squares method, have been used in this field. However, up to now, it has not been possible to say if one of these techniques gives the optimal solution of the problem to be solved. A genetic algorithm is introduced to search for the optimal optical thin film design. This paper discusses the problem of thin film design in greater detail. It shows how a genetic algorithm can evolve the design for better performance. Examples of designs obtained by GA optimization techniques are given.

An Automatic Analysis System for Firearm Identification Based on Ballistics Projectile

Characteristic markings on the cartridge and projectile of a bullet are produced when a gun is fired. Over thirty different features within these marks can be distinguished, which in combination produce a ”fingerprint” for identification of a firearm. Given a means of automatically analyzing features within such a firearm fingerprint, it will be possible to identify not only the type and model of a firearm, but also each individual weapon as effectively as human fingerprint identification can be achieved. In this paper, a new analytic system based on fast Fourier transform (FFT) for identifying the projectile specimens digitized using the line-scan imaging technique automatically is proposed. Experimental results show that the proposed system has the ability of efficient and precise analysis and identification for projectiles specimens.

Global optimization for optical thin-film design using Latin Squares

There are many advanced local and global optimization techniques, such as Gradient, Simplex, Flip-flop, Needle, Genetic and Simulated Annealing, which have been successfully applied to optical thin-film design. However, all these optimization techniques either require a selection of a reasonable starting design, which is a big obstacle to an inexperienced designer, or they have some kind of inbuilt random feature, which may give rise to different answers each time. To find the true global optimized solution for a thin film design problem, we need to solve an array of interlinked multi-dimensional simultaneous equations. Until recently, for more than just a few layers, this has been a very difficult task, requiring the use of a supercomputer and highly skilled programming. By using orthogonal Latin Square theory and an experimental design methodology in a search space reduction process, a Windows based program has been written that can operate on even a 20 MHz 386 computer. It can find the global optimum design for up to 23 layers using as many dispersive and lossy materials as one wishes, within a period of hours. Additionally this methodology (called DGL-Optimization) allows the use of multiple target spectra with such as both s & p polarization, for reflection and transmission simultaneously.

Directional filtering as a multi-purpose tool in image processing

Directional filtering is commonly known as a technique for selective linear enhancement. In addition to this traditional utility, we present in this paper some examples demonstrating that it can be used as a multi-purpose tool in image processing, by adjusting its parameters. These examples show that: (1) applying directional filtering in conjugate directions simultaneously can lead to the enhancement of specific conjugate linear features on an image, which is useful in interpretation of geophysical and Landsat images, satellite and aerial photographs that contain many conjugate linear features; (2) directional filtering can be used as illuminating sources to produce shaded relief images that show 3D effect; by combining the directional filtered result with its original image, this 3D effect becomes much more profound; (3) the directional filtered result can be weighted by different factors, which allows some subtle features on an image to be identified.

Web based image database system for ballistic firearm identification

Computerised ballistic identification system can be used to streamline the process and increase the efficiency and accuracy. Web base FireBall aims to provide such system via Internet. This web application development is a practical approach to extend the formal desktop Fireball system to web application by using web database and web programming mechanism. The goal of this research is to develop an integrated and secured web ballistic system, so that the historical data and images can be accessed anywhere in the world to support the ballistic firearm identification.

Application of modified conjugate directional filtering for remote sensing image processing

In image processing, enhancement of conjugate linear features on an image is required sometimes. This enhancement can be achieved by using two directional filters in two specific conjugate directions. Traditionally the filtered data from the two conjugate directions are presented in two separate images. This is inconvenient for making a direct comparison between linear features in the two conjugate directions. We recently proposed a new method named the modified conjugate directional filtering (MCDF) to combine two direction-filtered results in two conjugate directions into one image that retains the maximum information of the linear features in the two conjugate directions. Since conjugated features are often seen on remote sensing images, in this paper we present a few cases in using the MCDF to enhance the conjugated features on some remote sensing images. These examples show that the MCDF is a useful tool for linear feature enhancement not only in the two selected directions, but in other directions as well.

Application of fuzzy logic in seismic zonation

In this paper we propose a fuzzy logic system that uses crustal structural features (seismotectonics) and historic seismic activities (seismicity) as two fuzzy inputs for seismic zonation. Seismotectonics is a combination of features from the regional and deep geology, neotectonics, stress field and geophysics whereas seismicity is defined by historic earthquakes and their damages to some specific areas. Applying this fuzzy system to the northeastern Tibetan Plateau, a well-known intraplate seismic region in the world, outlines are not only the existing well-recognised seismic zones where large earthquakes took place in history, but also some areas where there have been no strong shocks occurred for the last 2000 years. The traditional statistical methods are not able to evaluate such areas due to the lack of the historic seismicity information.

Cleavage luminescence from silicon

This paper reports on further research into the structure and properties of the cleaved surfaces of silicon, using vacuum cleavage luminescence detection methods. The experiments involved detecting the luminescence produced by cleaving thin silicon plates within a high vacuum, by a process of converting the luminescence to an amplified electrical signal. The experiments were based on the assumption that surface cleavage and reconstruction may cause electrons to become excited, and the resulting recombination process which involves the combining of an electron in the conduction band with a hole in the valence band, would result in an emission of energy that would be detectable. This hypothesis was supported by simple calculations that predicted that every broken atomic bond on a silicon surface should radiate one photon, thus generating a strong detectable emission signal.