Yongwoon Choi

Measurement of a Container Crane Spreader Under Bad Weather Conditions by Image Restoration

Recently, anti-sway systems for container crane spreaders have played an important role in efficiently handling cargo at terminals. A camera-based measurement system for the anti-sway systems, which consists of two reference marks on the upper surface of the spreader and a camera mounted on the cranes trolley, has been developed. However, the quality of the reference marks in images taken in bad weather such as mist, fog, and rain is degraded, and consequently, the measurement accuracy also deteriorates. To preserve measurement accuracy in bad weather, a rapid method for image restoration is needed. We propose a method that restores the quality of reference marks in images by modeling the degradation of the light intensity. The two constants in the model are estimated from the variances and averages of the intensities in two reference mark images extracted from two images taken at different heights of the spreader. The height of the spreader, as measured by the system, is also fed back into the restoration. The measurement accuracy attained by the method under simulated bad weather conditions is compared with the accuracy achieved in normal conditions. It is found that measurement accuracy is preserved, with only small errors of up to 1.39 mm under a swaying motion. The processing time of the method is 0.83 ms on average, which does not affect the measurement period of the system.

Position measurement of container crane spreader using an image sensor system for anti-sway controllers

Speedy operation efficiency is constantly required of the container crane operators who work for cargo handling in the port. They are experts in suppressing the sway motion which is generated in the process of the handling. This undesirable motion may deteriorate the operation efficiency due to strained nerves of the operator. Therefore, an anti-sway controller has been required on practical job-site. To accomplish an adequate anti-sway control, sensors are also required to be able to measure the position of a crane spreader moving on. In this paper, we propose a position measurement system with an image sensor. The image sensor detects two devised landmarks attached to the upside surface of a spreader to measure the displacement and height to one. The image processing technique used for the sensor is a kind of robust template matching method, named vector code correlation (VCC), devised to consider the conditions of real environments. We employ the simple mechanism, so that the system can be easily applicable to existing container cranes. The details and performance of the measurement system are introduced together with the experimental results performed with the pilot system. We will also show that the system has the sufficient efficiency and accuracy for the use on practical job-site.

Measurement system design for sway motion based on image sensor

Speedy operation efficiency is constantly required of the container crane operators who work for cargo handling in the port. They are experts in suppressing the sway motion which is generated in the process of the handling. This undesirable motion may deteriorate the operation efficiency due to strained nerves of the operator. Therefore, an anti-sway controller has been required on practical job-site. To accomplish an adequate anti-sway control, sensors are also required to be able to measure the position of a crane spreader moving on. In this paper, we propose a position measurement system with an image sensor. The image sensor detects two devised landmarks attached to the upside surface of a spreader to measure the displacement and height to one. The image processing technique used for the sensor is a kind of robust template matching method, named Vector Code Correlation (VCC), devised to consider the conditions of real environments. We employ the simple mechanism, so that the system can be easily applicable to existing container cranes. The details and performance of the measurement system are introduced together with the experimental results performed with the pilot system. We will also show that the system has the sufficient efficiency and accuracy for the use on practical job-site.

Monocular 3D palm posture estimation based on feature-points robust against finger motion

The usability of wearable augmented reality (AR) systems would improve by letting users arbitrarily display virtual information on their palm and simultaneously manipulate it as tablet computers or smartphones. To realize such interaction between users and virtual information, we aim to robustly estimate 3-D palm posture against finger motion. This is based on the assumption that finger motion is separately estimated from palm posture and applied to manipulation of displayed information. In addition, the capability of electric sources, sensors, and processors are very limited in wearable computers. For this reason, by using a monocular camera and estimating palm posture from only a few image feature-points, we achieve an efficient estimation that satisfies real-time constraint on wearable computers. The accuracy and the robustness of our method are demonstrated by qualitative and quantitative comparison with a widely-used cardboard maker. Additionally, we confirmed that our method is run on a mobile computer at the average of 12.44 msec per frame.

Hardware design of vector code correlation method for high-speed template matching

Template matching has been applied to the image processing of the fields such as security and robot vision to recognize a target from images. The method used in those fields requires the high-speed processing and robustness to illumination change or occlusion. To accommodate this, we employ a method, named the vector code correlation (VCC) that codes the gradient of intensity change, but the method on software is not enough in aspect of image processing time. Therefore, the purpose of this study is to achieve the template matching in real-time by designing hardware to implement the VCC method at high speed. The hardware design consists of five modules doing pipeline processing and is able to realize in relatively small scale. In the present paper, experimental results of 60 fps obtained with the designed hardware and its configuration are shown.

Estimation for a hand position and orientation robust to motion of fingers using a web camera

We present a method to estimate the position and the orientation of a hand from a captured image for applications which need to manipulate instinctively and freely using a hand. We use a hand as a tool instead of an input device such as a mouse or a marker, and we use only a web camera instead of multiple-camera or an infrared camera, in order to improve mobility and usability. We focused on four valley-points between neighboring fingers as invariant feature points. By detecting four valley-points, the method makes possible the robust estimation for the position and the orientation of a hand regardless of the bending and the closing of fingers. In the method, we are able to move fingers freely when all valley-points are viewable. In an experiment, we qualitatively evaluated the robustness and usefulness of the proposed method by drawing 3DCG objects.

Measurement accuracy on indoor positioning system using spread spectrum ultrasonic waves

This paper describes significant factors on spatial error distribution measured in a given indoor space with a local positioning system which utilizes spread spectrum ultrasonic waves generated by narrow directional ultrasonic transmitters. The positioning error has interestingly been discussed from the viewpoints of the dilution of precision (DOP) and the ranging error due to the transmitter directional characteristic, mainly. The indoor 3-D position is calculated by means of an algorithm using distances from the transmitters located at the four corners of the ceiling to a mobile receiver. From the result of these experiments, it has been usefully found that the positioning error can be more largely influenced on the DOP of vertical direction than it of horizontal direction and that its average is of 3.43[cm] rms for all 3-D measurable points. These findings obtained from the work imply that the 3-D indoor positioning can be realized with an acceptable centimeter-order of precision. Moreover, they suggest that it has to be considered from the viewpoints of not only the ranging error, which could be yielded by the distance measurement, but also the DOP, which is inevitably obtained by a geometrical arrangement relative to the transmitters and receiver.

FPGA-Based Image Processor for Sensor Nodes in a Sensor Network

A field-programmable-gate-array- (FPGA-) based image processor which can be used for sensor nodes in a sensor network has been proposed and developed. Image processors for the nodes must satisfy requirements such as low power consumption, small circuitry scale, and modifiability of the hardware architecture. By developing an image proces- sor designed using an FPGA, SRAM modules, and the vector code correlation method which is suitable for the construc- tion of the target hardware architecture, it was possible to ensure that the processor satisfies these requirements. In this paper, we present the details of this image processor, which employs the template matching method for target tracking as well as the background subtraction method for object extraction. In addition, in order to verify its applicability in sensor nodes, we demonstrate the usefulness of the image processor from the results of an experiment in which the template matching and background subtraction methods were implemented simultaneously.

Cloud VR system with immersive interfaces to collect human gaze-controls and interpersonal-behaviors

In this study, we present a cloud VR system with immersive interfaces for collecting human gaze-controls and interpersonal-behaviors. Human beings can log in to a VR space and naturally communicate with each other and a robot by using immersive interfaces. An HMD (Head Mounted Display) device and a Kinect sensor provide immersive visualization and motion control, respectively. If users have these devices, they can attend experiments with human-robot interactions in the VR space from around the world. Human gaze-controls and interpersonal-behaviors are collected to database while human beings and a robot interact with each other. The proposed VR system enables to collect huge number of human gaze-controls and interpersonal-behaviors by using immersive interfaces. Application experiments to learn object attributes from human beings and to observe greetings by two persons demonstrate the effectiveness of the proposed VR system to collect human gaze-controls and interpersonal-behaviors.

A method for supporting robot's actions using virtual reality in a smart space

The purpose of this study is to construct a system that is able to supporting interactively the users such as human and robots in a smart space where sensors and actuators are dispersively installed. The virtual reality utilized as a platform in the system is built by representing to put the information together from sensors in the smart space where the users share the information and collaborate each other. The system has provided to support the task assignment and route planning for robots. The task assignment is to assign a suitable robot for executing the task by using the information acquired from the virtual reality. The route planning is a function to serve a safe path by pointing a sequent sub-goal in case of what happened to the sensors on a robot or the robot under complicated surroundings. The effectiveness of the system to support robots and experimental results implemented with two robots are presented in this paper