Currently, the common methods for image retrieval are content-based, while the abilities of image feature representation of these methods are very limited. In this paper, a new image retrieval method for binary images based on Deep Belief Networks (DBN) and Softmax classifier is proposed, which classifies the image data-set into some categories with the DBN and Softmax classifier first, and then classifies the query image in the same way, and those images in the same category will be returned as the similar images of the query image. Unlike the existing image retrieval models, the new method aims to provide a more effective representation and extraction measure by simulating the architecture of human visual system, and it is not necessary to set the threshold manually for this method like most of the existing methods based on the hamming distance computation. Experimental results show that the proposed method can get better recall and precision than some existing methods, such as perceptual hash algorithm and shape-based algorithm.
11 September 2024 in IEEE Internet of Things Journal
In this article, we address the key problem of sensing and communication integrated resource allocation for 6G-empowered distribution grid hierarchical coordinated control. First, we construct a novel information timeliness metric for electric semantic communication, namely, Peak Age of Semantics (PAoS), which covers the entire lifecycle of information sensing, semantic compression, semantic transmission, and semantic decoding. Second, we propose a sensing and semantic communication integrated resource allocation algorithm based on Top- $\text {N}^{2}$ and hybrid knowledge–statistic-driven fuzzy reinforcement learning. A deep fuzzy neural network is utilized to build a knowledge model between the grid operating state and decision making. The knowledge is embedded into statistic-driven model of reinforcement learning to enhance accuracy of upper confidence bound (UCB) utility evaluation. Finally, simulations based on realistic application scenarios indicate that compared with two comparison algorithms, the proposed algorithm reduces average PAoS by 4.72% and 9.49%, and the maximum PAoS by 5.76% and 13.57%. Additionally, its end-to-end delay trend and semantic packet decoding success rate align more closely with semantic importance.
Power Internet of Things devices deployed in smart parks have strict time synchronization requirements. However, due to the limited availability of 4-G/5-G, optical fiber, and satellite, how to achieve low-cost, reliable, and efficient time synchronization based on existing power line carrier (PLC) and two-way power frequency automatic communication system (TWACS) integrated multimodal communication has become a key challenge. Motivated by this, a cluster-based time synchronization network structure is proposed that seamlessly integrates PLC and TWACS, which provides an accurate synchronization model by comprehensively considering frequency offset, timestamp recorded error, timestamp jitter error, and transmission delay. Moreover, we incorporate digital phase-locked loop to compensate synchronization error by correcting frequency offset in clusters. The goal is to realize the minimization of the average time synchronization error within clusters. To achieve this goal, we introduce a coalition formation game-based dynamic clustering algorithm (CFG-DC), designed to effectively address the proposed optimization challenge. CFG-DC achieves dynamic clustering with consideration of time synchronization error awareness by performing leave-and-join decision iteratively to maximize synchronization error-based utility function. Simulation distinctly demonstrate the remarkable achievement of CFG-DC in attaining heightened levels of accuracy in time synchronization.
As the focal length of an optical lens in a conventional camera is limited, it is usually arduous to obtain an image in which each object is focused. This problem can be solved by multi-focus image fusion. In this paper, we propose an entirely new multi-focus image fusion method based on decision map and sparse representation (DMSR). First, we obtained a decision map by analyzing low-scale images with sparse representation, measuring the effective clarity level, and using spatial frequency methods to process uncertain areas. Subsequently, the transitional area around the focus boundary was determined by the decision map, and we implemented the transitional area fusion based on sparse representation. The experimental results show that the proposed method is superior to the other five fusion methods, both in terms of visual effect and quantitative evaluation.
With the explosive growth in the quantity of electrical equipment in distribution networks, traditional manual inspection struggles to achieve comprehensive coverage due to limited manpower and low efficiency. This has led to frequent equipment failures including partial discharge, insulation aging, and poor contact. These issues seriously compromise the safe and stable operation of distribution networks. Real-time monitoring and defect identification of their operation status are critical to ensuring the safety and stability of power systems. Currently, commonly used methods for defect identification in distribution network electrical equipment mainly rely on single-image or voiceprint data features. These methods lack consideration of the complementarity and interleaved nature between image and voiceprint features, resulting in reduced identification accuracy and reliability. To address the limitations of existing methods, this paper proposes distribution network electrical equipment defect identification based on multi-modal image voiceprint data fusion and channel interleaving. First, image and voiceprint feature models are constructed using two-dimensional principal component analysis (2DPCA) and the Mel scale, respectively. Multi-modal feature fusion is achieved using an improved transformer model that integrates intra-domain self-attention units and an inter-domain cross-attention mechanism. Second, an image and voiceprint multi-channel interleaving model is applied. It combines channel adaptability and confidence to dynamically adjust weights and generates defect identification results using a weighting approach based on output probability information content. Finally, simulation results show that, under the dataset size of 3300 samples, the proposed algorithm achieves a 8.96–33.27% improvement in defect recognition accuracy compared with baseline algorithms, and maintains an accuracy of over 86.5% even under 20% random noise interference by using improved transformer and multi-channel interleaving mechanism, verifying its advantages in accuracy and noise robustness.