The detection of banana fruits is an important part of intelligent management in the banana plantation. To detect the banana fruit quickly and accurately in the complex orchard environment, this article proposes a method based on the latest deep learning algorithm to detect the banana fruit. Using a monocular camera, we applied the YOLOv4 neural network algorithm to extract the deep features of banana fruits, realizing accurate detection of different banana sizes. The detection algorithm achieved a 99.29% detection rate, the average execution time was 0.171s, the shortest execution time was 0.135s, and the AP was 0.9995. Moreover, the detection results were discussed with the YOLOv3 algorithm and the machine learning algorithm. Compared with the machine learning algorithm, deep learning algorithm was superior to both detection accuracy and detection time. YOLOv4 had higher detection confidence and higher detection rate than YOLOv3. The results show that the proposed method could realize the fast detection of different varieties and different maturity in banana plantations, under different illumination and occlusion conditions, and provide information for banana picking, maturity and yield estimation.
The fast growth of surveillance video big data presents great challenges to the video coding technology. Most existing video coding techniques target for visual quality optimization, while the ultimate utility of surveillance videos mainly lies in intelligent analyses, e.g., pedestrian detection and vehicle tracking. In view of this, we aim at proposing an efficient, standard-compatible and simultaneously analysis-friendly coding framework for intelligent surveillance videos. In particular, the foreground objects are first extracted from the video sequence by accurate background modeling. Subsequently, the foregrounds can be constructed as a sequence and compressed in higher quality while very few background pictures are required to signal in lower quality. At the decoder side, the foreground objects can be directly used for efficient analysis tasks and the surveillance videos can be also reconstructed by synthesizing background and foreground frames. The effectiveness and potential of the proposed framework have been demonstrated in the pedestrian detection application, where the coding bits can be greatly saved with the detection accuracy being well maintained.
With the rapid development of optical waveguide technology, an essential component for creating compact and efficient AR devices, its application in the field of augmented reality (AR) is becoming more and more widespread. When paired with a variety of microdisplay technologies (LCoS, DLP, Micro‐LED), it enables the production of high‐performance AR products. In order to achieve a deeper immersive experience, designs such as large‐angle light‐emitting microdisplay technology and dilated pupil gratings are often used to provide a wider field of view and a larger eyebox (the viewing area where the eye can see the entire image). However, these designs pose challenges because they can exacerbate brightness and color inhomogeneity, disrupting the user's experience. We address this shortcoming by proposing an evaluation model based on the human visual system, which introduces a new method for constructing frequency‐domain filters in a uniform color space and simulates the processing of early human vision , thereby providing a more accurate reflection of human eye perception and a more realistic evaluation of AR display performance. By offering valuable insights and guidance, our model effectively enhances the quality of AR displays, positioning it as an indispensable tool for researchers, engineers, and industry professionals.
Hyperpolarized noble gas MRI is a new technique for imaging of gas spaces and tissues that have been hitherto difficult to image, making it a promising diagnostic tool. The unique properties of hyperpolarized species, particularly the non-renewability of the large non-equilibrium spin polarization, raises questions about the feasibility of hyperpolarized noble gas MRI methods. In this paper, the critical issue of T1 relaxation is discussed and it is shown that a substantial amount of polarization should reach the targets of interest for imaging. We analyse various pulse sequence designs, and point out that total scan times can be decreased so that they are comparable or shorter than tissue T1 values. Pulse sequences can be optimized to effectively utilize the non-renewable hyperpolarization, to enhance the SNR, and to eliminate image artifacts. Hyperpolarized noble gas MRI is concluded to be quite feasible.
Inter prediction serves as the foundation of prediction based hybrid video coding framework. The state-of-the-art video coding standards employ the reconstructed frames as the references, and the motion vectors which convey the relative position shift between the current block and the prediction block are explicitly signalled in the bitstream. In this paper, we propose a high efficient inter prediction scheme by introducing a new methodology based on virtual reference frame, which is effectively generated with the deep neural network such that the motion data does not need to be explicitly signalled. In particular, the high quality virtual reference frame is generated with the deep learning based frame rate up-conversion (FRUC) algorithm from two reconstructed bi-prediction frames. Subsequently, a novel CTU level coding mode termed as direct virtual reference frame (DVRF) mode, is proposed to adaptively compensate for the current to-be-coded block in the sense of rate-distortion optimization (RDO). The proposed scheme is integrated into the HM-16.6 software, and experimental results demonstrate significant superiority of the proposed method, which provides more than 3% coding gains on average for HEVC test sequences.
Tobacco plants recognizing and counting accurately is very important in the tobacco plant management. In this paper, we propose an algorithm for tobacco plants recognizing and counting which consists of four main steps: first, we apply the unmanned aircraft to acquire the tobacco images. Second, the tobacco image is converted into the Lab color space, and then b channel in Lab space is processed based on morphological reconstruction. Third, the candidate regions which might contain tobacco plants are extracted based on the processed b channel. Finally, SVM (Support Vector Machine) is employed to classify the candidate regions as tobacco plants or not. The proposed method has been evaluated on a tobacco image data set. Experimental results (96.1% of accuracy and 94.3% of sensitivity) showed that the proposed method is feasible.
In the brain Magnetic Resonance (MR) images, the boundary of each encephalic tissue is highly irregular. It is difficult to accurately detect the encephalic tissues. Owing to its powerful capacity in solving non-linearity problems, Support Vector Machine (SVM) has been widely used in object detection. The conventional SVMs, however, assume that each feature of a sample has the same importance degree for the detection result, which is not a true representation of real applications. In addition, the parameters of the SVM and its kernel function also affect detection result. In this study, Immune Algorithm (IA) was introduced in searching for the optimal feature weights and the parameters simultaneously. An Immune Feature Weighted SVM (IFWSVM) method was used to detect encephalic tissues in MR images. Theoretical analysis and experimental results showed that the IFWSVM has better performance than the conventional methods.
When the region of interest (ROI) is smaller than the object, one can increase MRI speed by reducing the imaging field of view (FOV). However, when such an approach is used, features outside the reduced FOV will alias into the reduced-FOV image along the phase-encoding direction. Reduced-FOV methods are designed to correct this aliasing problem. In the present study, we propose a combination of two different approaches to reduce the acquired FOV: 1) two-dimensional (2D) spatially-selective RF excitation, and 2) the unaliasing by Fourier-encoding the overlaps using the temporal dimension (UNFOLD) technique. While 2D spatially-selective RF excitation can restrict the spins excited within a reduced FOV, the UNFOLD technique can help to eliminate any residual aliased signals and thus relaxes the requirement for a long RF excitation pulse. This hybrid method was implemented for MR-based temperature mapping, and resulted in artifact-free images with a fourfold improvement in temporal resolution.
The new rural construction refers to the construction of the economic, political, cultural and social aspects of rural areas under the socialist system, in accordance with the requirements of the new era, and finally realizes the rural construction into economic prosperity, perfect facilities, beautiful environment and civilization the goal of the new socialist countryside.Jiao Zhou Yangko as the most extensive basis of the rural mass sports project, in the construction of a socialist harmonious society and new rural construction, to find an opportunity for their own development, and this paper is in this context through the overview of Jiao Zhou Yangko, Giving its new features, and put forward the corresponding countermeasures.Enrich the spiritual and cultural life of the masses in the construction of new countryside, and provide practical guiding significance for the better sustainable development of Jiao Zhou Yangko.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)
