Neural Network Method for Colorimetry Calibration of Video Cameras

To transfer the color data from a device (video camera) dependent color space into a device? independent color space, a multilayer feedforward network with the error backpropagation (BP) learning rule, was regarded as a nonlinear transformer realizing the mapping from the RGB color space to CIELAB color space. A variety of mapping accuracy were obtained with different network structures. BP neural networks can provide a satisfactory mapping accuracy in the field of color space transformation for video cameras.

Is It OK to Install Video Cameras in Taxicabs?

Wuhu City in Anhui Province initiated a program to install GPS video surveillance and alarm systems in taxicabs across the city in August.It is projected that all of the city’s 3,024 cabs will be equipped with security cameras by mid-October.Wuhu is not alone in adopting such a system.Several other Chinese cities,including Chongqing,Xi’an and Nanchang,have engaged in a similar campaign.

Multimodal background model with noise and shadow suppression for moving object detection

A statistical multimodal background model was described for moving object detection in video surveillance. The solution to some of the problems such as illumination changes, initialization of model with moving objects, and shadows suppression was provided. The background samples were chosen by thresholding inter-frame differences, and the Gaussian kernel density estimation was used to estimate the probability density function of background intensity. Pixel's neighbor information was considered to remove noise due to camera jitter and small motion in the scene. The hue-max-min-diff color information was used to detect and suppress moving cast shadows. The effectiveness of the proposed method in the foreground segmentation was demonstrated in the traffic surveillance application.

Spark-generated bubble collapse near or inside a circular aperture and the ensuing vortex ring and droplet formation

This paper aims to study the oscillation of a sparkgenerated submerged bubble located near or inside a circular aperture made in a flat plate using high-speed visualization technique. In the case of a bubble oscillating near an aperture the initial free surface of the water was set at the bottom surface of the plate. The effects of aperture size and bubblefree surface distance on the bubble behavior as well as on the ensuing droplet dynamics are investigated. It was found that the direction of the bubble reentrant jet was towards the aperture or away from it respectively when the normalized aperture size was smaller or greater than a certain critical value. In addition, a toroidal vortex ring was observed to form, which rotated inwards as it moved away from the aperture. It was also found that if the bubble was incepted at a distance sufficiently away from a supercritical size aperture a single droplet could be produced. In the case of a bub- ble initiated in the middle of a circular aperture submerged just beneath the water free surface, the bubble was found to take the shape of an ellipsoid during its expansion. Then a reentrant jet was initiated and pierced the bubble from its top side.

Experimental Investigation on Cavitating Flow Shedding over an Axisymmetric Blunt Body

Nowadays,most researchers focus on the cavity shedding mechanisms of unsteady cavitating flows over different objects,such as 2D/3D hydrofoils,venturi-type section,axisymmetric bodies with different headforms,and so on.But few of them pay attention to the differences of cavity shedding modality under different cavitation numbers in unsteady cavitating flows over the same object.In the present study,two kinds of shedding patterns are investigated experimentally.A high speed camera system is used to observe the cavitating flows over an axisymmetric blunt body and the velocity fields are measured by a particle image velocimetry（PIV）technique in a water tunnel for different cavitation conditions.The U-type cavitating vortex shedding is observed in unsteady cavitating flows.When the cavitation number is 0.7,there is a large scale cavity rolling up and shedding,which cause the instability and dramatic fluctuation of the flows,while at cavitation number of 0.6,the detached cavities can be conjunct with the attached part to induce the break-off behavior again at the tail of the attached cavity,as a result,the final shedding is in the form of small scale cavity and keeps a relatively steady flow field.It is also found that the interaction between the re-entrant flow and the attached cavity plays an important role in the unsteady cavity shedding modality.When the attached cavity scale is insufficient to overcome the re-entrant flow,it deserves the large cavity rolling up and shedding just as that at cavitation number of 0.7.Otherwise,the re-entrant flow is defeated by large enough cavity to induce the cavity-combined process and small scale cavity vortexes shedding just as that of the cavitation number of0.6.This research shows the details of two different cavity shedding modalities which is worthful and meaningful for the further study of unsteady cavitation.

Influence of longitudinal magnetic field on metal transfer in MIG arc welding

Metal transfer is an important phenomenon in metal inert gas （MIG） arc welding with longitudinal alternating magnetic field. It is of great significance to observe the metal transfer modes under different excitatory currents and frequencies. However, it is very difficult to view the metal transfer process directly during welding, due to the strong interference from the arc light. To obtain the relationship between the metal transfer modes and the different magnetic fields, a high-speed video camera was used to acquire the images of globules. Different metal transfer modes under the conditions of different magnetic fields and welding parameters were studied. The experiment shows clear images of droplet transfer as well as influence of longitudinal magnetic field on both metal transfer and globule shape.

Rapid Texture Mapping from Image Sequences for Building Geometry Model

An effective approach, mapping the texture for building model based on the digital photogrammetric theory, is proposed. The easily-acquired image sequences from digital video camera on helicopter are used as texture resource, and the correspondence between the space edge in building geometry model and its line feature in image sequences is determined semi-automatically. The experimental results in production of three-dimensional data for car navigation show us an attractive future both in efficiency and effect.

A robust system for real-time pedestrian detection and tracking

A real-time pedestrian detection and tracking system using a single video camera was developed to monitor pedestrians. This system contained six modules: video flow capture, pre-processing, movement detection, shadow removal, tracking, and object classification. The Gaussian mixture model was utilized to extract the moving object from an image sequence segmented by the mean-shift technique in the pre-processing module. Shadow removal was used to alleviate the negative impact of the shadow to the detected objects. A model-free method was adopted to identify pedestrians. The maximum and minimum integration methods were developed to integrate multiple cues into the mean-shift algorithm and the initial tracking iteration with the competent integrated probability distribution map for object tracking. A simple but effective algorithm was proposed to handle full occlusion cases. The system was tested using real traffic videos from different sites. The results of the test confirm that the system is reliable and has an overall accuracy of over 85%.

Simulation and Calibration of Hydro-Debris 2D Model （HD2DM） to Predict the Particle Segregation Processes in Debris Flow

The principal aim of a vertical two-dimensional numerical model development is for estimating the particle tracing and mechanism of 10 mm and 2.5 mm debris. The particle tracing movement can be visually analyzed by using a high speed video camera （HSVC）. A numerical model was developed using the Marker and Cell Method, which involves a Subgrid-Scale （SGS） model and the Particle Source in Cell （PSI-Cell） Method. The transportation processes of debris and air bubble were simulated in lagrangian form by introducing air bubbles and debris markers. Air bubble movement characteristics were simulated by this numerical model. Bigger particles flow at the upper part, while smaller particles attach near to the bottom. This phenomenon is similar to what we observed in the experimental studies. As a conclusion, the calibration processes for velocity was successful. The value of virtual mass （CM） was found to be one of the most important criteria that should be considered in the calibration process, as this parameter dominates fundamental characteristics of sediment particle movement in the lagrangian numerical scheme. The best fitted CM in this study was 0.35. The mean average velocity value ranging from 1.2% to 22.61% is obtained from the velocity results of numerical studies compared to the experimental studies.

Use of Accelerators for Biomechanical Analysis of Walking Motion Aided by Wheeled Walking Frames

Walkers improve self-reliability. We examined the effectiveness of a newly developed wheeled walking frame for use by physically handicapped persons. Unstable gaits in walker users were analyzed by tri-axial accelerometers and a motion capture system. Several markers were placed on subjects＇ backs and legs. Subjects were requested to walk around a test course at a comfortable speed, while their motion was recorded by two high-speed video cameras. The activities performed on the test course comprised standing, normal walking, fast walking, and walking over a barrier. Any accidental falls were also recorded. We established the characteristic rules of gait motion using a walker. Furthermore, we demonstrated that gait characteristics are more conveniently extracted from acceleration sensors than from motion capture systems, since the sensors can be affixed to subjects for self-monitoring and goal achievements. The methods employing acceleration sensors are considered suitable for determining the average gait motions of elderly persons living in nursing homes, and can be used to evaluate walking motion before and aider rehabilitation.

Evaluation of Skills in Swing Technique in Classical Japanese Swordsmanship in laido Using Sensors

In this study, we analyzed the swing motions of more experienced practitioner and new practitioner of iaido players by using tri-axial acceleration sensor and gyro sensor. Iaido is a modern Japanese martial art/sport. In this way, the acceleration and gyro sensor measurement enabled detailed motion information at the installation points to be displayed in a short time, thus making it possible to easily extract the objective problems. Although it was not possible to confirm by the acceleration and angular velocity measurements the detailed motion of the entire body as obtained in the 2D motion analysis with a high-speed camera, it was confirmed that the acceleration and gyro sensor is an evaluation means that can be installed easily and can provide the exercise information in a short time as an objective index.

An approach for ergonomic design of mouse wheel

A new method for ergonomic design of a computer mouse is proposed in this paper. In the method, the movements of joints and tip of the forefinger during operating a mouse was captured by a high-speed video camera. The captured videos were ana- lyzed and an algorithm was developed to decide the size and location of the mouse wheel according to ergonomic principles. The al- gorithm was then coded in a software package with Visual C＋＋ and OpenGL languages. Results of the calculation and simulation agreed well with those of the experiments. The software can also be used for shape design of mouse body, buttons and their layouts.

Design and performance of a video-based laser beam automatic alignment system

A laser alignment system is applied to a high power laser facility for inertial confinement fusion. A design of the automated, close-loop laser beam alignment system is described. Its function is to sense beam alignment errors in a laser beam transport system and automatically steer mirrors preceding the sensor location as required to maintain beam alignment. The laser beam is sampled by a sensor package, which uses video cameras to sense pointing and centering errors. The camera outputs are fed to a personal computer, which includes video digitizers and uses image storage and software to sense the centroid of the image. Signals are sent through the computer to a stepper motor controller, which drives stepper motors on mirror mounts preceding the beam sampling location to return the beam alignment to the prescribed condition. Its optical principles and key techniques are given. The pointing and centering sensitivities of the beam alignment sensor package are analyzed. The system has been verified on the

Estimating color appearance of pearlescent bottles using digital camera

Pearlescent coatings become highly popular to the modern generation of color rendering materials due to their unique color effect. However, it is quite challenging to evaluate its color appearance by traditional color measurement. A low-cost camera is a highly efficient device for multi-geometry color appearance estimation for pearlescent bottles, which has been achieved through the camera characterization, sample image capturing, and then mathematical transforming from RGB （red, green, and blue） values to color appearance attributes based on the color appearance model of CIECAM02. A tele-spectroradiometer for physical measurement together with visual assessment is applied for comparison with the camera method to evaluate the accuracy of camera predictions and discuss the applicability of CIECAM02. The experimental results indicate that the camera data have strong correlation with the physical measurement and also fit well with visual data except for a slight slope shift existing in lightness due to a divinable psychophysical magnitude variation for spatial-dependent color samples. Hence it is feasible to estimate the color appearance of pearlescent bottles using a digital camera.

The mechanism of penetration increase in A-TIG welding

The mechanism of the increasing of A-TIG welding penetration is studied by using the activating flux we developed for stainless steel, The effect of fluxon the flow and temperature fields of weld pool is simulated by the PHOENICS software, It shows that without flux, the fluid flow will be outward along the surface of the weld pool and then down, resulting in a flatter weld pool shape. With the flux, the oxygen, which changes the temperature dependence of surface tehsion grads froma negative value toa positive value, can cause significant changes onthe Weld penetration. Fluid flow will be inward along the surface of the weld pool toward the center and then down. This fluid flow pattern efficiently transfers heat to the weld root and produces a relatively deep and narrow weld.This change is the main cause of penetration increase. Moreover, arc construction can cause the weld width to become narrower and the penetration to become deeper, but this is not the main cause of penetration increase. The effects Of flux on fluid flow of the weld pool surface and arc profiles were observed in conventiOnai TIG welding and in A-TIG welding by using high-speed video camera; The fluid flow behavior was visualized in realtime scale by micro focused X-ray transmission video observation system. The result indicated that stronger inward fluid flow patterns leading to weld beads with narrower width and deeper penetration could be apparently identified in the case of A-TIG welding.The flux couldchange the direction of fluid flow i-n welding pool. It has a good agreement with the simulation result.