Improved estimation of defocus blur and spatial shifts in spatial domain: a homotopy-based approach

This paper presents a homotopy-based algorithm for the recovery of depth cues in the spatial domain. The algorithm specifically deals with defocus blur and spatial shifts, that is 2D motion, stereo disparities and/or zooming disparities. These cues are estimated from two images of the same scene acquired by a camera evolving in time and/or space. We show that they can be simultaneously computed by resolving a system of equations using a homotopy method. The proposed algorithm is tested using synthetic and real images. The results confirm that the use of a homotopy method leads to a dense and accurate estimation of depth cues. This approach has been integrated into an application for relief estimation from remotely sensed images.     

Exploration trees on highly complex scenes: A new approach for 3D segmentation

A new strategy for automatic object extraction in highly complex scenes is presented in this paper. The method proposed gives a solution for 3D segmentation avoiding most restrictions imposed in other techniques. Thus, our technique is applicable on unstructured 3D information (i.e. cloud of points), with a single view of the scene, scenes consisting of several objects where contact, occlusion and shadows are allowed, objects with uniform intensity/texture and without restrictions of shape, pose or location. In order to have a fast segmentation stopping criteria, the number of objects in the scene is taken as input. The method is based on a new distributed segmentation technique that explores the 3D data by establishing a set of suitable observation directions. For each exploration viewpoint, a strategy [3D data]-[2D projected data]-[2D segmentation]-[3D segmented data] is accomplished. It can be said that this strategy is different from current 3D segmentation strategies. This method has been successfully tested in our lab on a set of real complex scenes. The results of these experiments, conclusions and future improvements are also shown in the paper.     

Fault tolerant approach for biosignal-based robot control

This paper proposes a fault tolerant framework for biosignal-based robot control with multiple sensor electrodes. In this approach, to cope with sensor faults, a reliable joint torque estimation model is selected from a group of estimation models based on sensor failure classifiers. The correlation among the electromyography (EMG) signal streams is used as input feature vectors for fault detection. To validate our proposed method, we artificially disconnect an EMG electrode or detach one side of an EMG probe from the skin surface during elbow-joint torque estimation experiments with five participants. When one EMG sensor electrode experiences one of the problems, the experimental results show that the joint torque can be estimated with significantly fewer errors using our proposed approach than a joint torque estimation method without sensor fault detection or than a method with a conventional sensor fault detection algorithm. Furthermore, we controlled a mannequin-arm-attached one-DOF exoskeleton based on the estimated torque profiles by generating movements with the estimated torque derived from the selected model.     

A noncausal approach for PID control

A new approach for the improvement of the set-point following performances achieved by a PID controller is presented. Basically, it consists of applying a suitable command signal to the closed-loop control system in order to achieve a desired transient response when the process output is required to assume a new value. This command signal is determined by means of a stable input–output procedure for which a closed-form solution is presented, thus making the technique suitable to implement in the industrial context. Simulation and experimental results show that high performances are obtained despite the presence of model uncertainties and, above all, almost independently on the PID tuning. Thus, the PID gains can be selected in order to guarantee good load rejection performances without impairing the set-point transient responses.     

Toward designing intelligent PDEs for computer vision: An optimal control approach

Many computer vision and image processing problems can be posed as solving partial differential equations (PDEs). However, designing a PDE system usually requires high mathematical skills and good insight into the problems. In this paper, we consider designing PDEs for various problems arising in computer vision and image processing in a lazy manner: learning PDEs from training data via an optimal control approach. We first propose a general intelligent PDE system which holds the basic translational and rotational invariance rule for most vision problems. By introducing a PDE-constrained optimal control framework, it is possible to use the training data resulting from multiple ways (ground truth, results from other methods, and manual results from humans) to learn PDEs for different computer vision tasks. The proposed optimal control based training framework aims at learning a PDE-based regressor to approximate the unknown (and usually nonlinear) mapping of different vision tasks. The experimental results show that the learnt PDEs can solve different vision problems reasonably well. In particular, we can obtain PDEs not only for problems that traditional PDEs work well but also for problems that PDE-based methods have never been tried before, due to the difficulty in describing those problems in a mathematical way.     

A general fuzzy-statistical clustering approach for estimating the time of change in variable sampling control charts

Despite their capability in monitoring the variability of the processes, control charts are not effective tools for identifying the real time of such changes. Identifying the real time of the change in a process is recognized as change-point estimation problem. Most of the change-point models in the literature are limited to fixed sampling control charts which are only a special case of more effective charts known as variable sampling charts. In this paper, we develop a general fuzzy-statistical clustering approach for estimating change-points in different types of control charts with either fixed or variable sampling strategy. For this purpose, we devise and evaluate a new similarity measure based on the definition of operation characteristics and power functions. We also develop and examine a new objective function and discuss its relation with maximum-likelihood estimator. Finally, we conduct extensive simulation studies to evaluate the performance of the proposed approach for different types of control charts with different sampling strategies.     

Laplacian pyramid networks: A new approach for multispectral pansharpening

Pansharpening is about fusing a high spatial resolution panchromatic image with a simultaneously acquired multispectral image with lower spatial resolution. In this paper, we propose a Laplacian pyramid pansharpening network architecture for accurately fusing a high spatial resolution panchromatic image and a low spatial resolution multispectral image, aiming at getting a higher spatial resolution multispectral image. The proposed architecture considers three aspects. First, we use the Laplacian pyramid method whose blur kernels are designed according to the sensors’ modulation transfer functions to separate the images into multiple scales for fully exploiting the crucial spatial information at different spatial scales. Second, we develop a fusion convolutional neural network (FCNN) for each scale, combining them to form the final multi-scale network architecture. Specifically, we use recursive layers for the FCNN to share parameters across and within pyramid levels, thus significantly reducing the network parameters. Third, a total loss consisting of multiple across-scale loss functions is employed for training, yielding higher accuracy. Extensive experimental results based on quantitative and qualitative assessments exploiting benchmarking datasets demonstrate that the proposed architecture outperforms state-of-the-art pansharpening methods. Code is available at https://github.com/ChengJin-git/LPPN.     

A multi-plane approach for text segmentation of complex document images

This study presents a new method, namely the multi-plane segmentation approach, for segmenting and extracting textual objects from various real-life complex document images. The proposed multi-plane segmentation approach first decomposes the document image into distinct object planes to extract and separate homogeneous objects including textual regions of interest, non-text objects such as graphics and pictures, and background textures. This process consists of two stages—localized histogram multilevel thresholding and multi-plane region matching and assembling. Then a text extraction procedure is applied on the resultant planes to detect and extract textual objects with different characteristics in the respective planes. The proposed approach processes document images regionally and adaptively according to their respective local features. Hence detailed characteristics of the extracted textual objects, particularly small characters with thin strokes, as well as gradational illuminations of characters, can be well-preserved. Moreover, this way also allows background objects with uneven, gradational, and sharp variations in contrast, illumination, and texture to be handled easily and well. Experimental results on real-life complex document images demonstrate that the proposed approach is effective in extracting textual objects with various illuminations, sizes, and font styles from various types of complex document images.     

Launch control for sport motorcycles: A clutch-based approach

Launch control is an innovative active system which offers an appealing automatic feature for sport motorcycles. Due to the high power of the engine and to the complexity of the torque transfer dynamics driving a smooth and safe start from a standstill might be difficult for non-expert riders in some working conditions. As such, an active controller capable of safely managing the vehicle acceleration from a standing start might effectively support the rider as far as both safety and performance are concerned. To control such a maneuver, two possible actuators can be employed: the engine throttle and the clutch. This paper illustrates the analysis and the design of a clutch-based launch controller for sport motorcycles, demonstrating that it allows the accurate tracking of an acceleration set-point thanks to a dedicated clutch position controller. Experimental results are presented that demonstrate the effectiveness of the proposed approach.     

Road detection algorithm for Autonomous Navigation Systems based on dark channel prior and vanishing point in complex road scenes

Vision-based road extraction is essentially important in many fields, such as for intelligent traffic and robot navigation. However, the road detection in urban or ill-structured roads is still very challenging at current stage, and the existing methods often suffer from high computational complexity. This paper reports a novel and efficient method for road detection in challenging scenes. First, the dark channel based image segmentation is proposed to distinguish a rough road region from complex background noise, which is envisioned to reduce the workload of road detection. Furthermore, instead of using the conventional pixel-wise soft voting, a new voting strategy based on the vanishing point and the properties of the segmented regions is proposed to further reduce the computation time of road extraction stage. Finally, the segmented region which has the maximum voting value is selected as the road region. Experimental results demonstrated that the proposed algorithm shows superior performance in different kinds of road scenes. It can remove the interference from pedestrians, vehicles and other obstacles. Our method is about 40 times faster in detection speed, when compared to a recently well-known approach.