Search

In the context of cattle farm environments, intricate environmental interferences have presented challenges that impede seamless data acquisition. This paper introduces a novel approach, the integration of a fusion gate transformation mechanism and a variational autoencoder GAN, which we term GCT-VAE-GAN, aimed at enhancing low-light images from cattle farm settings. Initially, our approach involves the design of an encoding network tasked with augmenting the original images. Subsequently, we advance our methodology by formulating a generative network to effectively address the challenges of image diversification and poor image quality. Notably, the inclusion of an attention mechanism block within the FFN layer facilitates the fusion of these extracted features, resulting in the generation of high-quality images. Furthermore, to achieve proficient image discrimination, we implement a dual-discriminator structured discriminative network for the conclusive image discrimination task. The culmination of our approach involves the formulation of a comprehensive joint loss function, thereby constituting the core of the model’s loss module. Moreover, through comparative experiments, we aptly demonstrate the remarkable superiority of the GCT-VAE-GAN approach. The conducted experiments reveal the model’s consistent performance and resilience under varying illumination scenarios. The outcomes of our study underscore its significant relevance in elevating the quality of low-light images within cattle farm contexts. Furthermore, our approach exhibits the potential to enhance the efficacy of computer vision tasks, signifying a notable stride toward improved agricultural imaging techniques.

Cattle’s individual identification plays a crucial role in effectively managing large farms. To enhance agricultural efficiency, promote the digital transformation of animal husbandry, and improve animal welfare, it is essential to employ advanced identification technologies capable of real-time monitoring of cattle individual. This paper introduces a novel network called Open Pose Mask R-CNN (OP-Mask R-CNN) for individual cattle identification, which combines Open Pose with the Mask R-CNN network. Three key strategies are presented to improve the identification of individual cattle. First, optimize the number of convolutional layers in the Mask R-CNN backbone network, i.e., ResNet101. Second, introduce an Open Pose-based bovine skeleton feature extraction method. Finally, construct a fusion mechanism that combines the attention module, the convolutional block attention module (CBAM), the open pose module, and the ResNet101. Experimental results demonstrate that our proposed method achieves a 5.6% increase in recognition accuracy and improves recognition speed compared to the original Mask R-CNN model. This work strikes a balance between accuracy and complexity, facilitating the development of a lightweight bovine individual recognition technique.

Theoretical estimates based on both offline calculation and an online model are made to study the wavenumber variations during wave refraction in a marginal ice zone (MIZ). Analysis of in situ observations from the MIZ of the Arctic Ocean confirms the conclusion drawn based on theoretical estimates as well as the simulation ability of our new model to describe wave evolution in the MIZ. In wave refraction, variation of wavenumber in magnitude is determined by the incident wavenumber and the ice mass. A larger incident wavenumber or a larger ice mass leads to a larger variation of wavenumber in magnitude, though during refraction, the variation of wavenumber in magnitude is more sensitive to the incident wavenumber because it is proportional to the cube of the incident wavenumber, while it is linearly proportional to the ice mass. On the other hand, both the angle of refraction and the deflection are determined by the angle of incidence, the incident wavenumber and the ice mass. A larger angle of incidence, a shorter wave or larger ice mass results in a larger angle of refraction and a larger deflection. All these conclusions are supported by offline calculations, an online model and in situ observations. The online model also suggests that the instrument should be kept at a distance from the ice edge to obtain more general information of the open ocean.

In this paper, cooperative cognitive radio networks are considered, in which a primary user (PU) and an off-the-grid secondary user (SU) co-exist by exploiting simultaneous wireless information and power transfer. Based on a two-phase relaying model, adaptive power splitting is performed at the SU for information decoding while collecting the energy remaining in the first phase. The energy harvested is then used to forward the decoded primary signal, with the secondary signal superimposed in the second phase. To enhance the utilization of both the spectrum and energy, an energy-aware retransmission approach is proposed for enabling successful decoding at the SU while collecting a reasonable amount of energy for relaying. The outage probability and throughput are theoretically analyzed for both the PU and the SU. To provide more analytical insights, tight performance upper and lower bounds are obtained in closed forms. Our results demonstrate that a mutually beneficial relationship can be built between the PU and the SU under proper parameter configurations. Furthermore, a performance tradeoff with respect to the number of retransmissions is demonstrated, where additional performance gains can be achieved by the proposed retransmission approach under unfavorable conditions of high rate, low power, and weak channels.

In the isolated power system of renewable energy power generation inverter and diesel-engine generator operating in parallel, the generator usually runs in the voltage and rotate speed droop control mode, and the three-phase inverter runs in the constant power-control mode. In this case, a sudden load change will be completely undertaken by the generator, which may lead to its overload or reverse power. In order to solve this problem, this paper has proposed an emergency control strategy based on the detection of the rate of system frequency change. In emergency conditions, the strategy can be used to automatically adjust the inverter output power to deal with the aforementioned overload or reverse power and prevent the system from shedding load and avoid protective actions of the generator. Thus, the power-supply capacity and continuity of the system are guaranteed, and its safety and stability are also improved. Time-domain simulations and experiments show that the proposed strategy is feasible and effective.

Multichannel cognitive radio (CR) has been proposed to improve system throughput by selecting multiple idle subchannels for transmissions. However, these idle subchannels are often noncontinuous, which cannot be utilized by the traditional access models, such as orthogonal frequency division multiplex and multicarrier code division multiple access. In this paper, a CR system based on transform domain communication system has been designed to access the non-continuous spectrum, which detects the spectrum status by energy detection and marks the spectrum availability as a spectrum marker vector. A basis carrier generated from the spectrum marker vector may concentrate the power on the idle subchannels. The bit error rate of the CR system is analyzed both in the cases of spectrum inconsistency and multiple access. The transmission data is modulated on the basis carrier using binary modulation and cyclic code shift keying modulation. A system optimization unit is designed to improve the system throughput by jointly optimizing sensing threshold and transmission power of each subchannel. The simulation results have shown that the proposed CR system may outperform the spread spectrum system and achieve larger throughput through performance optimization.

In cognitive radio (CR), the detection performance of traditional spectrum sensing may be decreased due to the multipath fading and shadowing effect. However, unmanned aerial vehicle (UAV) can receive a higher strength signal without suffering severe fading and shadowing. In this paper, an UAV-based CR is proposed to improve spectrum sensing performance and access the idle spectrum. The UAV performs circular flight with the PU locating at the center, and the flight period is divided into sensing radian and transmission radian. Only when the absence of the PU has been detected within sensing radian can the UAV access the spectrum. The spectrum sensing optimization is formulated as an optimization problem, which seeks to maximize the effective throughput of the UAV by optimizing the sensing radian subject to the constraint of the interference throughput. Then, virtual cooperative spectrum sensing is proposed to improve sensing performance in fading channel, and the effective throughput is maximized by jointly optimizing local sensing radian and the number of sensing slots. The simulation results show that better UAV transmission performance can be obtained by optimizing spectrum sensing, and the virtual cooperative spectrum sensing outperforms the single spectrum sensing especially in deep fading channel.