Many image segmentation methods have been proposed over the last few decades. However, the results of segmentation are usually evaluated only visually, qualitatively, or indirectly by the effectiveness of the segmenta...
详细信息
ISBN:
(纸本)9781424433520
Many image segmentation methods have been proposed over the last few decades. However, the results of segmentation are usually evaluated only visually, qualitatively, or indirectly by the effectiveness of the segmentation on the subsequent processing steps. Such methods are either subjective or tied to particular applications. A few quantitative evaluation methods have been proposed, but these early methods have been based entirely on empirical analysis and have no theoretical grounding. In this paper, we put forward the concept of segmentation entropy, based on which a novel objective segmentation evaluation method is proposed. This method uses region entropy as the basis for measuring the uniformity of pixel characteristics within segmented regions and segmented entropy as a whole measure to provide a relative qualitative score that can be used to compare both various parameterizations of one particular segmentation method as well as fundamentally different segmentation techniques.
Accurately extrinsic camera self-calibration namely the positions and orientations of the networked cameras by themselves, is essential for many applications such as surveillance, intelligent environment and traffic m...
详细信息
ISBN:
(纸本)9781424479351
Accurately extrinsic camera self-calibration namely the positions and orientations of the networked cameras by themselves, is essential for many applications such as surveillance, intelligent environment and traffic monitoring. This paper describes an efficient, range-free and anchor-free method for self-calibration the extrinsic parameters of the cameras in a non-overlapping camera sensor networks. The proposed method is based on the method proposed by Ali on the 2004's CVPR. Knowledge of the locations or angles, got from the assisted sensor (accelerometer or angular-accelerometer) installed on the moving object provide additional effective constraints on the optimization problem in order to compute the cameras, poses. Simulation result show that the iteration times, the calibration error, the volume of the data needed by the improved method are far less than the original method. The advantage of the method is that it can be applied even when the target sharp turns out of any camera's Field of view (FoV) with little steps.
Traditionally, fish farming involves manual methods for fish identification. However, with the advancement of computer vision technology, an automatic identification system can be developed. This study introduces a fi...
详细信息
ISBN:
(纸本)9798350360875;9798350360868
Traditionally, fish farming involves manual methods for fish identification. However, with the advancement of computer vision technology, an automatic identification system can be developed. This study introduces a fish identification method based on detecting fish body pose using a lightweight convolutional neural network model. The Micropterus salmoides was used as the study object to construct a fish pose dataset. The target detection frame has also been updated to a rotating format to better recognize fish in all orientations. Feature vectors were constructed based on the pose estimation, and the random forest algorithm was applied for feature degradation and fish identity identification. The fish pose detection network achieved an accuracy rate of 0.98 on the test set. Detection-based tracking was also used to track the trajectory of fish targets. This method can provide a foundation for fish individual identification, size detection, and precision breeding devices, optimizing efficiency in fish farming.
Farming robots need a fast and robust image segmentation module to apply targeted treatments, which require the ability to distinguish, in real time, between crop and weeds. Existing solutions make use of visual class...
详细信息
ISBN:
(纸本)9781728171401
Farming robots need a fast and robust image segmentation module to apply targeted treatments, which require the ability to distinguish, in real time, between crop and weeds. Existing solutions make use of visual classifiers that are trained on large annotated datasets. However, generating large datasets with pixel-wise annotations is an extremely time-consuming task. In this work, we tackle the crop/weed segmentation problem by using a synthetic image generation method to augment the training dataset without the need of manually labelling the images. The proposed approach consists in training a Generative Adversarial Network (GAN), which can automatically generate realistic agricultural scenes. As a difference with respect to common GAN approaches, where the network learns how to reproduce an entire scene, we generate only instances of the objects of interest in the scene, namely crops. This allows to build a generative model that is more compact and easier to train. The generated objects are then placed into real images of agricultural datasets, thus creating new images that can be used for training. To evaluate the performance of the proposed approach, quantitative experiments have been carried out using different segmentation network architectures, showing that our method well generalizes across multiple architectures.
Big Data technology has demonstrated its effectiveness in several areas, and several projects confirm the existence of innovative opportunities. However, The security and privacy issues are gradually magnified in thes...
详细信息
Switched reluctance motors (SRM) have a cheap and robust structure, which makes them interesting for electric vehicle applications. In this paper, a new optimization scheme for the design of SRM is proposed. The desig...
详细信息
Sliding Mode control (SMC) technique has been used in an extensive manner in many practical applications especially in motion control systems. This paper investigates non-linear discrete time systems accommodating inp...
详细信息
ISBN:
(纸本)9781509002108
Sliding Mode control (SMC) technique has been used in an extensive manner in many practical applications especially in motion control systems. This paper investigates non-linear discrete time systems accommodating input delay. Firstly, input delay is removed by introducing a smith predictor that converts the original discrete system into delayed free version of the system and makes it solvable. Then, for an effective control "reaching law method" is used to design control law and construction of sliding surface for the delayed free system. The Chebyshev Neural Networks (CNNs) are used to approximate the unknown non-linearity. Simulation shows the robustness of the control scheme.
The no. 2 reversing mill installed in the cold rolling plant of the Wakayama Steel Works is described. It has been built with the aim of ensuring high-precision strip gauge and flatness. It includes a fully digitized ...
详细信息
The no. 2 reversing mill installed in the cold rolling plant of the Wakayama Steel Works is described. It has been built with the aim of ensuring high-precision strip gauge and flatness. It includes a fully digitized AC variable-speed drive system in the electric control and advanced technology that improves gauge accuracy. The specifications and features of the mill are set forth, the characteristics of the electrical equipment are discussed, and the gauge-control system is examined in more detail.
This paper presents the analysis of a brushless doubly fed reluctance machine (BDFRM) for renewable energy applications with an emphasis on the rotor design. The BDFRM stator includes two sets of windings, i.e., the c...
详细信息
This paper presents the analysis of a brushless doubly fed reluctance machine (BDFRM) for renewable energy applications with an emphasis on the rotor design. The BDFRM stator includes two sets of windings, i.e., the control winding (CW) and power winding (PW). The PW is connected to the grid, while the CW links the grid by a power converter of fractional rating to regulate the operating speed. This reduces the converter cost, and the brushless configuration enhances the reliability. In this paper, the reluctance rotor of the BDFRM, which is similar to that of a synchronous reluctance machine, is designed and analyzed in order to improve the performance. The influence of the total MMF and its allocation to the two stator windings are also discussed in relation to the machine performance. Finite-element analysis is used for performance simulation. The experimental study validates the proposed design for a small BDFRM.
The paper addresses the problem of sensor location regarding the solvability of the disturbance decoupling problem in finite automata. Two methods are given to find the measured output, which guarantees the solvabilit...
详细信息
暂无评论