The complementary advantages of point cloud and image can provide more accurate 3D and semantic information to the model. Aiming at the problems that most existing methods adopt a single fusion strategy and thus fail to achieve deep fusion of image and point cloud features, this paper studies and analyzes the existing fusion strategy of image and point cloud data, and proposes a model based on the fusion of projected point cloud and image features. The model utilizes a projection fusion and feature fusion strategy, introduces a wide threshold processing in the projection module, meanwhile applies the fusion of point clouds and image features after projection cropping, finally integrates both features in depth by adding a weight fusion layer in the feature fusion stage. Extensive experiments on the public KITTI dataset demonstrate that mAP of the proposed method is improved by 3.34% in the average values of easy difficulty compared with similar models, indicating that the algorithm is more effective in 3D object detection with point cloud and image fusion.

Performance analysis of cloud computing server provides the basis for ensuring Quality of Service (QoS), and the service strategy of server will directly affect the analysis of performance indicators. The performance indicators of QoS are usually defined in the form of Service Layer Agreement (SLA), such as the average response time, the average queue length, immediate service probability and so on. In this work, Service performance analysis models based on Geo/G/1 queuing system and queuing system with the vacation of the server are proposed. In these models, we analyze the main performance indicators of cloud computing server for the different parameters: the time between arrive of the task, the time of service, and the time of the provision of vacation. Furthermore, we discuss the optimizing concurrent number of the cloud computing.

The study of sign language involves the intersection of many fields and disciplines. At present, the two mainstream research directions of sign language recognition are data gloves [1] and visual sign language recognition [2] . The former uses the data collected by the sensor for sign language recognition and translation, while the latter uses the camera to capture the user's hand characteristics for sign language recognition and translation. In this paper, an improved convolutional neural network (CNN) [3] and long short-term memory(LSTM) [4] neural network combined sign language recognition system, which is different from the current only for sign language recognition and translation, but also for sign language generation function is designed. For the first time, this system uses a PyQt designed GUI interface. Once in the system, users can select sign language recognition and translation capabilities, capture images via OpenCV, and then use the trained CNN neural network for special processing. The model can then identify American sign language through LSTM decisions. The user can also click the voice button, the system will be based on the user's voice to convert the corresponding gesture image into the same pixels, and write to the video file. Experimental results show that sign language recognition rate is 95.52% compared with similar algorithms [5], and sign language [6] (American sign language and Arabic numerals) is 90.3%.

The image intelligent processing analysis technology uses a computer to imitate and execute some intellectual functions of the human brain, and realizes an image processing system with artificial intelligence, that is, an image processing analysis technology is an understanding of an image. The degree of intelligent automated analysis and processing is low, many operations need to be done manually, causing human error, inaccurate detection, and time-consuming and laborious. Deep learning method can extract features step by step in the original image from the bottom to the top. Therefore, based on feature analysis technology, this paper uses the deep learning method to intelligently and automatically analyse the visual image. This method only needs to send the image into the system, and then the manual analysis is not needed, and the analysis result of the final image can be obtained. The process is completely intelligent and automatically processed. First, improve the deep learning model and use massive image data to choose and optimize parameters. Results indicate that our method not only automatically derives the semantic information of the image, but also accurately understands the image accurately and improve the work efficiency.

The irrelevant background information in the personalized image is easy to be quantified into the same word as the main target, and the quantization process will inevitably cause the loss of a lot of visual information. This phenomenon will seriously reduce the quality of the generated theme when the personalized image content is complex. This paper proposes a Multi-Source Big Data Fusion Annotation (MSBDFA) model. The model obtains similar personalized images by analyzing the relevant multi-source information of the personalized images, and uses the annotations of the similar personalized images to label the personalized images. For the personalized images with complex background visual information, the personalized image retrieval based on complete information modeling uses the high-dimensional Gaussian distribution to directly model the continuous visual features of the personalized images, and uses the two-level spectral clustering algorithm to distribute the regional topics, so as to embed the complete local information contained in the visual features into the global features of the personalized image. Therefore, this method can completely retain visual information during the modeling process, so that the targets buried in the complex background can be better classified. The experimental results on the standard database show that the method proposed in this paper can generate high-quality personalized image subjects in complex scenes and has good retrieval performance.