No abstract

No abstract

No abstract

In this paper, we introduce an development device which can measure the working radius of crane truck, automatically. This device is equipped with a laser range finder, a CCD camera and two AC servo motors. Each motor has an encoder. First, we propose the method of object recognition to measure the position of mark set of measurement object, and secondly construct the tracking control system to the development device. Then, it has PI control system with an extended Kalman filter. Through experiments, we show the effectiveness of our proposed method.

In this paper, we develop the observation device to measure the working radius of a crane truck (shown in figure 1). The working radius is a distance between the center of rotation of a crane and the crane hook. The observation device is equipped with a CCD camera which can change its zoom, a laser range finder and two AC motors which have encoder. To measure the working radius, this paper propose two methods, the recognition method for a targer object and tracking control method for a moving crane hook.As for the recognition method, it is difficult to recognize a crane hook, because there are various shapes in a crane hook. Then we attach the cross mark which is shown in figure 2 to a crane hook to simplify recognition of a target object. Using image processing, we recognize the cross mark. In this paper, recognition means detection of the cross shape. Before that, we detect circles on the image to reduce the process time. Recognition of this paper consists of binarization, smooting, labeling, circle detection and cross detection, and each process is devised to shorten processing time and improve a success rate of the recognition.As for the tracking control, a mark which is attached to crane hook is moved by a operator, so we have to control posture of the observation device in order to track to the moving mark. When we control the observation device based on the image information, tracking delay occure because enoumous time is need to image capturing and image processing. Then we realize the tracking control without delay by using extended Kalman filter. Additionally we control a zoom in order to keep the mark size in the constant on the image to the mark which moved in the depth direction.Position of the mark is calculated by information of the observation device which are the image of the camera, distance information of the laser range finder and posture of the observation device by encoders. Through some experiments, we verify the availability of the observation device. First, we verify performance of recognition and tracking. In this experiment, we arrange the overhead crane is attached the cross mark 5 meters ahead from the observation device (shown in figure 3). Then, we move the mark. The movement of the mark are pendulum movement and movement to the backward. Using the observation device, we observe the orbit of the moving mork. From the observed result which is shown in figure 4, we can confirm that the observation device can observe a moving mark stably.Further, we verify the measurement accuracy using real crane truck. In this experiment, we use the observation device which we developed and the total station which is current product, and compare these measurement result. In this paper, we suppose that the result of the total station is accurate. As a result of measuring the working radius of a crane, we can confirm that the observation device can measure the working radius of a crane to an accuracy of 50 mm. It follows from results of experiments that the observation device which we developed is use...