In this paper, we propose a fast labeling algorithm based on block-based concepts. Because the number of memory access points directly affects the time consumption of the labeling algorithms, the aim of the proposed algorithm is to minimize neighborhood operations. Our algorithm utilizes a block-based view and correlates a raster scan to select the necessary pixels generated by a block-based scan mask. We analyze the advantages of a sequential raster scan for the block-based scan mask, and integrate the block-connected relationships using two different procedures with binary decision trees to reduce unnecessary memory access. This greatly simplifies the pixel locations of the block-based scan mask. Furthermore, our algorithm significantly reduces the number of leaf nodes and depth levels required in the binary decision tree. We analyze the labeling performance of the proposed algorithm alongside that of other labeling algorithms using high-resolution images and foreground images. The experimental results from synthetic and real image datasets demonstrate that the proposed algorithm is faster than other methods.
Different from other references that use Ziegler–Nichols method or another trial and error method to design the proportional–integral–derivative (PID) controller, this research is based on the precise controller design in combination with the root locus criterion (RL). According to the system properties and the task requirements, we applied the model reduction method, root locus criterion, root contour (RC), pole-zero placement, frequency response, and orthodox PID criterion to design various controllers. Besides, we analyzed the system responses in a time domain to obtain the optimal controller parameters. We also designed a robust controller to give the system better relative stability through the comparisons of system performances in the frequency domain. Furthermore, in this research, we also created the “Reverse Gain PID Controller” to deal with the negative effects of the non-minimum phase system. Not only does this controller improve both the transient response and steady-state error but also enhance the relative stability of the system at the same time. Finally, according to the system properties, this research can make the controllability of UAV much better by a variety of design methods and flexible controller applications, and then expand the application fields and scopes of UAVs. In sum, the research content is academic, practical, and very efficient as it can be used as reference material that can be applied in further follow-up research.
Object segmentation and object labeling are important techniques in the field of image processing. Because object segmentation techniques developed using two-dimensional images may cause segmentation errors for overlapping objects, this paper proposes a three-dimensional object segmentation and labeling algorithm that combines the segmentation and labeling functions using contour and distance information for static images. The proposed algorithm can segment and label the object without relying on the dynamic information of consecutive images and without obtaining the characteristics of the segmented objects in advance. The algorithm can also effectively segment and label complex overlapping objects and estimate the object’s distance and size according to the labeling contour information. In this paper, a self-made image capture system is developed to capture test images and the actual distance and size of the objects are also measured using measuring tools. The measured data is used as a reference for the estimated data of the proposed algorithm. The experimental results show that the proposed algorithm can effectively segment and label the complex overlapping objects, obtain the estimated distance and size of each object, and satisfy the detection requirements of objects at a long-range in outdoor scenes.