Body Height Measurements in Images

Abstract:  For any measurement of a person's height in an image, a difference exists between the actual height of the person and the image measurement. In order to gain knowledge about statistical behavior of differences between actual and measured heights it is necessary to make reference recordings, e.g., of test persons under the same recording conditions. To test whether the differences are dependent on camera and further circumstances, an experiment was set up which involved the measurement of 22 test persons using three cameras of varying quality. Reproducibility of measurements per image appears to be strongly dependent on the camera (quality), whereas systematic bias differs with the view point of the camera. Operator dependency of the measurement process is found, so its repetition by different operators is recommended.     

Comparison of the Performance of Two Methods for Height Estimation

Abstract: In the case study, two methods of performing body height measurements in images are compared based on projective geometry and 3D modeling of the crime scene. Accuracy and stability of height estimations are tested using reconstruction images of test persons of known height. Given unchanged camera settings, predictions of both methods are accurate. However, as the camera had been moved in the case, new vanishing points and camera matches had to be created for the reconstruction images. 3D modeling still yielded accurate and stable estimations. Projective geometry produced incorrect predictions for test persons and unstable intervals for questioned persons. The latter is probably caused by the straight lines in the field of view being hard to discern. With the quality of material presented, which is representative for our case practice, using vanishing points may thus yield unstable results. The results underline the importance of performing validation experiments in casework.     

Generating a Corpus of Mobile Forensic Images for Masquerading user Experimentation

The Periodic Mobile Forensics (PMF) system investigates user behavior on mobile devices. It applies forensic techniques to an enterprise mobile infrastructure, utilizing an on‐device agent named TractorBeam. The agent collects changed storage locations for later acquisition, reconstruction, and analysis. TractorBeam provides its data to an enterprise infrastructure that consists of a cloud‐based queuing service, relational database, and analytical framework for running forensic processes. During a 3‐month experiment with Purdue University, TractorBeam was utilized in a simulated operational setting across 34 users to evaluate techniques to identify masquerading users (i.e., users other than the intended device user). The research team surmises that all masqueraders are undesirable to an enterprise, even when a masquerader lacks malicious intent. The PMF system reconstructed 821 forensic images, extracted one million audit events, and accurately detected masqueraders. Evaluation revealed that developed methods reduced storage requirements 50‐fold. This paper describes the PMF architecture, performance of TractorBeam throughout the protocol, and results of the masquerading user analysis.     

A Novel Pixel‐Based Method to Estimate the Instantaneous Velocity of a Vehicle from CCTV Images

In traffic forensic identifications, it was usually unfavorable that estimated velocities of questioned vehicles derived from frame‐based methods eventually turned out to be close to but a bit less than speed limits. In this paper, a novel pixel‐based method was presented to estimate vehicles’ relatively instantaneous velocities. First, two adjacent pixels’ actual distance was acquired from the closed‐circuit television (CCTV) images. Then, the instantaneous average velocity within this one frame interval time (FIT) was calculated out. The frame‐based method and its evaluating result were also introduced in this paper. The results showed that the velocity estimated by this new pixel‐based method was higher than which derived from the frame‐based method. Employing this presented novel method in these cases, it could be more accurately identified whether or not the vehicles had exceeded the speed limits, and then the responsibilities could be consequently assigned impartially.     

Estimating the preappearance interval from temperature in Creophilus maxillosus L. (Coleoptera: Staphylinidae)

Abstract: Some carrion insects appear on carcasses late in decomposition. While using these late‐arriving taxa to estimate postmortem interval, it is vital to accurately estimate an interval preceding their appearance on a cadaver called the preappearance interval (PAI). This paper tests the possibilities of a PAI estimation from temperature in case of late‐arriving, predatory carrion beetle Creophilus maxillosus. The relationship between temperature and PAI of adult and larval C. maxillosus was modeled using the results of a large‐scale experiment on succession in forests of Central Europe. In both stages, PAI was found to be strongly, inversely related to temperature averaged for the duration of PAI. The PAI estimation from mean degree‐day accumulations (mDDA), estimated DDA, and “PAI*T” functions were tested using external successional data. The greatest accuracy of PAI estimates was achieved using an exponential “PAI*T” function.     

Estimating Projectile Perpendicular Impact Velocity on Metal Sheet Targets from the Shape of the Target Hole

Abstract:  The correlation between bullet hole shapes in metal and projectile impact velocity was examined. A series of shots were fired from an M-16A1 assault rifle of 5.56 mm caliber toward a 1-mm thick metal target. All shots were fired at a perpendicular angle to the metal sheets, and the velocity was measured just before the projectile hit the target. Velocities ranged between 400 and 900 m/sec. From the replica of the shooting hole, a perpendicular plane was created, showing the symmetrical properties of the hole. The best mathematical equation describing the shape of the entrance hole was the exponential function in the form:
The empirical equation of the hole defined using the regression method is:
This equation describes the general shape of shooting holes created by velocities ranging from 440 to 750 m/sec. From this equation, one can estimate the bullet velocity when it hits the target.     

Proficiency Testing as a Basis for Estimating Uncertainty of Measurement: Application to Forensic Alcohol and Toxicology Quantitations

Abstract: While forensic laboratories will soon be required to estimate uncertainties of measurement for those quantitations reported to the end users of the information, the procedures for estimating this have been little discussed in the forensic literature. This article illustrates how proficiency test results provide the basis for estimating uncertainties in three instances: (i) For breath alcohol analyzers the interlaboratory precision is taken as a direct measure of uncertainty. This approach applies when the number of proficiency tests is small. (ii) For blood alcohol, the uncertainty is calculated from the differences between the laboratory’s proficiency testing results and the mean quantitations determined by the participants; this approach applies when the laboratory has participated in a large number of tests. (iii) For toxicology, either of these approaches is useful for estimating comparability between laboratories, but not for estimating absolute accuracy. It is seen that data from proficiency tests enable estimates of uncertainty that are empirical, simple, thorough, and applicable to a wide range of concentrations.     

Source Camera Identification for Heavily JPEG Compressed Low Resolution Still Images*

Abstract: In this research, we examined whether fixed pattern noise or more specifically Photo Response Non‐Uniformity (PRNU) can be used to identify the source camera of heavily JPEG compressed digital photographs of resolution 640 × 480 pixels. We extracted PRNU patterns from both reference and questioned images using a two‐dimensional Gaussian filter and compared these patterns by calculating the correlation coefficient between them. Both the closed and open‐set problems were addressed, leading the problems in the closed set to high accuracies for 83% for single images and 100% for around 20 simultaneously identified questioned images. The correct source camera was chosen from a set of 38 cameras of four different types. For the open‐set problem, decision levels were obtained for several numbers of simultaneously identified questioned images. The corresponding false rejection rates were unsatisfactory for single images but improved for simultaneous identification of multiple images.     

Determination of Vehicle Speed from Recorded Video Using Reverse Projection Photogrammetry and File Metadata

The prevalence of security and in‐car video has increased the number of motor vehicle accidents captured on digital video. However, inconsistencies in how to accurately determine time and distance for vehicle speed has led to examinations with varying results. A potential solution for calculating time intervals is to use frame timing contained within many digital video file's metadata, recorded with 0.000001 sec precision. This paper examines a fatal motor vehicle accident where frame timing information was used with distance measurements from reverse projection photogrammetry to calculate vehicle speed. A margin of error was then calculated based on the accuracy in performing reverse projection photogrammetry distance measurements. The resulting speed calculation was then compared to event data recorder data and found to be within an average of ±1.43538 MPH. Using specific time intervals may lead investigators to more accurate speed calculations, specifically those involving variable frame rate video.     

Age Estimation from Radiographic Images of the Knee

This study examines the reliability of age estimation utilizing the Pyle and Hoerr atlas in relation to a modern Scottish population. The knee radiographs of 442 individuals (168 females, 274 males) were age assessed using the Pyle and Hoerr atlas. Analysis showed that there was a strong correlation between chronological age and estimated age (females R² = 0.968, males R² = 0.952). For females, the atlas method was most accurate between the ages of 9 and 15 years of age with an underage of 2.27 months and an overage of 2.38 months. For males, the atlas consistently overestimated age from the age of 9 years to the age of 16 years from between 0.14 and 8.81 months. The standard deviation for females was 9.86 months and for males was 10.75 months. This study showed that the Pyle and Hoerr atlas can be applied to a modern population with small modifications.     

Comparison of Frequentist Methods for Estimating the Total Weight of Consignments of Drugs

Abstract: A topic in forensic statistics is the estimation of the total weight of consignments of drugs based on subsamples of which a certain fraction may not contain drugs at all. The frequentist approach to this concentrates on obtaining confidence intervals for the total weight, based on estimation of the fraction of drugs and the mean and variance of the weights of drug units. The current study shows that the resulting confidence intervals are basically unreliable, since they are based on an underestimation of the variation of the underlying statistical process. Two alternatives are given that yield asymptotically correct results. These are not reliable for small subsamples either, though, because of the inherent multimodal behaviour of the sample mean. In cases where a relatively large fraction of the consignment contains no drugs, the confidence intervals reported in the literature should not be used in practice.     

Variation in the Measurement of Cranial Volume and Surface Area Using 3D Laser Scanning Technology

Abstract: Three‐dimensional (3D) laser scanner models of human crania can be used for forensic facial reconstruction, and for obtaining craniometric data useful for estimating age, sex, and population affinity of unidentified human remains. However, the use of computer‐generated measurements in a casework setting requires the measurement precision to be known. Here, we assess the repeatability and precision of cranial volume and surface area measurements using 3D laser scanner models created by different operators using different protocols for collecting and processing data. We report intraobserver measurement errors of 0.2% and interobserver errors of 2% of the total area and volume values, suggesting that observer‐related errors do not pose major obstacles for sharing, combining, or comparing such measurements. Nevertheless, as no standardized procedure exists for area or volume measurements from 3D models, it is imperative to report the scanning and postscanning protocols employed when such measurements are conducted in a forensic setting.     

Quantifying Uncertainty in Estimations of the Total Weight of Drugs in Groups of Complex Matrices

In case practice at forensic drug departments, multiple items from one seizure are sometimes sent in with the question: what is the total amount of drugs in the seizure? This may be complicated especially if impregnated material is involved such as clothes or rubber. Measurement uncertainty is typically stable on drug percentages, not weights, and subsampling may take place. It is recognized more and more that determination and reporting of uncertainty on estimators are an essential part of obtaining scientifically sound results in the forensic field. Methodology is described to quantify uncertainty on estimations of the total drug weight in groups of complex matrices, given simple statistical models, along a subdivision of five types of cases. Given each of these types, case examples are given where uncertainty is quantified in estimations of drug weights, by means of confidence intervals. The described models are statistically sound and relatively easy to implement.     

The Effects of High Dynamic Range on Fingerprint Images Processed by Photoshop

High dynamic range (HDR) imaging is a function that combines five images with different exposures into a single image. This technique may provide fine ridge details of fingerprint images for forensic latent fingerprint examination. Therefore, viewing fingerprints under optimal conditions is of paramount importance. This paper analyzes HDR and non‐HDR photos by using the Michelson contrast formula. The Michelson formula will provide a measurement to determine whether better contrast between the background and print can be achieved using the HDR function and if the background color affects the quality of the images. Two hypothesis were tested: (i) the HDR image provides more details of fingerprints with a better tone, greater clarity, and contrast than a normally exposed image regardless of the background color; (ii) the background color does not affect the quality of HDR fingerprint images overall, but the multi‐color background may increase the contrast of HDR fingerprint images in some cases.     

Upper‐Bounding the Incidence Rate of Associations Between Camouflage Uniforms and Surveillance Images*

Abstract: Camouflage garments can be associated with surveillance images of a crime scene even in the absence of unique wear marks or very high‐quality images. However, the probability of an accidental association, or incidence rate, is significant. The present work describes and validates a method for estimating the incidence rate based on a statistical model of the garment manufacturing process. The model was developed primarily for use with the current U.S. Army Combat Uniform (ACU), but can be applied to any camouflage garment. Eight garment manufacturers were studied, and all sources of variation in the manufacturing process were characterized. The marking and spreading procedures were found to be dominant and consistent sources of variation. However, some sources of variation, in particular those because of human operators, were not consistent enough to accurately characterize. Sources of variation that could not be well‐characterized were ignored in the statistical model, yielding a worst‐case estimate that is an upper‐bound to the true incidence rate. The model was evaluated for a variety of cases. Depending on the quality of the surveillance image, the manufacturing parameters, and the local population, incidence rates range from about 3% to negligibly small. The model was validated by returning to one manufacturer, and sampling a large number of completed garments and estimating empirical match probabilities. The empirical probabilities validated the estimates of the worst‐case incidence rate and also demonstrated that typical incidence rates are significantly lower.     

Establishing Likelihood Ratios for Patterned Garment Comparisons from Seam Measurement Data,

It is often challenging to ascribe an objective measure of confidence for identifications based on surveillance imagery from a crime scene. The present work seeks to address this deficiency in the case of garment comparison evidence by developing a quantitative method for establishing a conservative lower bound on the likelihood ratio (LR) for identifications involving patterned garments. The method is based on statistical analysis of pattern offset measurements taken from a sample of garments of the same type (manufacturer, style, and size) as the seized evidence. The developed analysis framework was demonstrated on different types of garments over a range of modeled surveillance imaging scenarios with variable image quality; the lower bounds on the LRs ranged from approximately 10–1 to over 400–1. The statistical model was tested and validated through a large‐scale empirical study involving both simulated and human observer‐performed garment comparisons.     

Comparing the Scoring of Human Decomposition from Digital Images to Scoring Using On‐site Observations

When in forensic casework or empirical research in‐person assessment of human decomposition is not possible, the sensible substitution is color photographic images. To date, no research has confirmed the utility of color photographic images as a proxy for in situ observation of the level of decomposition. Sixteen observers scored photographs of 13 human cadavers in varying decomposition stages (PMI 2–186 days) using the Total Body Score system (total n = 929 observations). The on‐site TBS was compared with recorded observations from digital color images using a paired samples t‐test. The average difference between on‐site and photographic observations was ?0.20 (t = ?1.679, df = 928, p = 0.094). Individually, only two observers, both students with <1 year of experience, demonstrated TBS statistically significantly different than the on‐site value, suggesting that with experience, observations of human decomposition based on digital images can be substituted for assessments based on observation of the corpse in situ , when necessary.     

Estimating Sex Using Metric Analysis of the Scapula by Postmortem Computed Tomography

Postmortem computed tomography (CT) has been extensively used in the last decade for identification purposes and in various anthropologic studies. Postmortem CT measurements of scapulae, analyzed using logistic discriminant function developed in this study, showed 94.5% accuracy in estimating sex. Data analyzed using the Dabbs and Moore‐Jansen (2010) discriminant function and the discriminant function generated in this study provided nearly identical results with disagreement in only one case. Height and weight were not statically significant in sex prediction. The results of this study show that data obtained from volume rendered postmortem CT images can be considered reliable and treated as a practical option to standard anthropological methods, especially in mass fatalities as a rapid triage tool for sex determination.     

A Test of the DSP Sexing Method on CT Images from a Modern French Sample

The hip bone is considered to be one of the most reliable indicators in sex determination. The aim of this study was to test the reliability of the DSP method for the hip bone proposed by Murail et al. (Bull Mem Soc Anthropol Paris, 17, 2005, 167) on a sample from a present‐day population in France (52 males and 54 females). Ten linear measurements were collected from three‐dimensional models derived from computed tomography images (CTI). To quantify the proportions of correct sex determinations, a more rigorous posterior probability threshold of 0.95 was applied. Using all 10 measurements, 92.3% of males and 97.2% of females were sexed correctly. The percentage of undetermined specimens varied depending on the used combination of measurements; however, all sexes were assigned with a 100% accuracy. This study proves that DSP is an appropriate and reliable tool for sex determination, based on dimensions obtained from CTI.