Search

Unsupervised salient image generation without the aid of prior assumptions has many applications in computer vision. We present three unique real-time saliency generation algorithms that provide state-of-the-art performance for greyscale and colour images. Our fastest method run under 50 ms per frame on average. Our algorithm introduces a novel weighted histogram of orientation feature to supplement image intensity for monochromatic image manifold ranking. We also provide a method of dimensional reduction for the non-normalized optimal affinity matrix (OAM) using principal components analysis; this novel technique allows faster computation and stabilization of the OAM inversion process. We compare our methods with 18 traditional and recent techniques using three standard and custom datasets including ECSSD, DUT-OMRON and MSRA10K totalling 32,536 images for colour and greyscale variations. The results show our method to be more than $$10{\times }$$ faster than the RC and GMR models and having similar or better precision performances.

This paper studies an image descriptor that mimics the retina's photo receiving cell pattern. Various pattern differencing combinations and second order differencing techniques were explored. This new method shows higher precision and recall performance than the classical BRIEF and steered BRIEF descriptors.

This study investigates the rotational invariance of various image features extracted from monocular thermal and photo camera images that can be used for robot navigation. In addition to the conventional SIFT, SURF, ORB, AKAZE, and BRIEF image features, a new non-Gaussian binary feature descriptor is introduced using Levy-Laplace probability density functions. The rotational, viewpoint and noise invariance of all features are compared.

In this paper, a nonlinear direct adaptive output control methodology is developed to address the problem of spacecraft attitude control under large dynamics uncertainties. The main advantage of the proposed approach over other adaptive approaches for spacecraft attitude control is that it can handle significantly large inertia uncertainties, without requiring any on-line estimation of the unknown moments of inertia. Furthermore, the implementation of the controller does not require a mathematical model of the plant as the control gain adaptation mechanism relies on feedforward signals from an ideal model designed to provide a satisfactory response to the desired attitude commands, as well as on output tracking errors between the uncertain spacecraft and the ideal model. By modeling the spacecraft as a square nonlinear state-space plant through the use of the modified Rodrigues parameters allows the system to satisfy the almost striclty passive conditions, which are required to establish the formal proof of stability. The performance of the new adaptive attitude control approach is illustrated in numerical simulations for both a simple rigid-body rest-to-rest maneuver and a high-fidelity ISS free-flyer robotic capture maneuver.

This paper presents a two-level place names identification method based on N-shortest path and Conditional Random Fields(CRFs) aiming at solving the low recall rate problem in Chinese place names identification. First, the rough segmentation method based on N-shortest path is used to improve the recall rate of Chinese place names identification at low level; Second, the result of rough segmentation is submitted to high level as one of the features of high-level place names identification. High level’ s CRFs model uses the feature which submitted by low level, single and complex features of place names words to tag the text. Adding the complex feature is conducive to mine the context information and improve accuracy rate of place names identification, and the result of text tagging could be combined with rules to identify place names finally. This two-level model ensures a high recall rate and improves the accuracy rate. During experiment, choose the mature corpora of People’ s Daily in January 1998 as training samples, which include 3128 place names(except duplicate names), and extract articles of People’ s Daily in 2003 randomly to carry out the test. Experimental results achieve a high recall rate, and this method is proved to be practical and effective.

An iterative algorithm can be modeled by a cyclic data-flow graph. The bottleneck for scheduling cyclic data-flow graphs lies on dependencies that form cycles. The maximum computation-time-to-delay ratio among all the cycles in the graph, gives a lower bound on pipeline schedule length. In this paper, a framework for algebraic transformations is proposed to reduce the lower bound. A novel algorithm is proposed to apply transformations within iterations or over iteration boundaries. A set of beneficial transformations are chosen, and applied simultaneously in each pass of the algorithm. A measure of criticalness on loops is used to identify transformations leading to potential lower bound reduction. Experimental results show that substantial reductions on the lower bound are achieved, and shorter pipelined schedules are generated.

Multiprocesssor systems have emerged as an important computing means for real-time applications and have received increasing attention than before. However until now, little research has been done on the problem of on-line scheduling of parallel tasks with deadlines in partitionable multiprocessor systems. In this paper, we propose a new on-line scheduling algorithm, called Deferred Earliest Deadline First (DEDF) for such systems. The main idea of the DEDF algorithm is to defer the scheduling as late as possible, so that a set of jobs is scheduled at a time instead of one at a time. For processor allocation using DEDF, we have introduced a new concept-Available Time Window (ATW). By using ATW the system utilization can be improved and thereby enabling the system to meet the deadline of more number of tasks. Simulation results for a hypercube indicate that the DEDF algorithm performs significantly better than the earlier proposed Buddy/RT and Stacking algorithms for a wide range of work loads.

Loop pipelining can be applied to a cyclic data-flow graph to reduce the iteration bound, which is the maximum computation-time-to-delay ratio among all the cycles in the data flow graph. Algebraic transformations can reduce the iteration bound substantially. However, resource constrained algebraic transformations for loop pipelining remains a hard problem because of the inherent nature of loop pipelining. In this paper, we propose a new method based on distribution graphs to solve this problem. A novel algorithm for algebraic transformation with resource constraints is provided, which works for non-pipelined schedules as well. Experimental results show that our algorithm is promising.