The raw trajectories contain large amounts of redundant data that bring challenges to storage, transmission and processing. Trajectory compression algorithms can reduce the number of positioning points while minimizing the loss of information. This paper proposes a heading maintaining oriented trajectory compression algorithm, which takes into account both position information and direction information. By setting an angle threshold, the algorithm can achieve a more accurate approximation of trajectories than traditional position-preserving trajectory compression algorithms. The experimental results show that the algorithm can ensure certain effect on the direction information and is more flexible.

The traditional text similarity measurement methods based on word frequency vector ignore the semantic relationships between words, which has become the obstacle to text similarity calculation, together with the high-dimensionality and sparsity of document vector. To address the problems, the improved singular value decomposition is used to reduce dimensionality and remove noises of the text representation model. The optimal number of singular values is analyzed and the semantic relevance between words can be calculated in constructed semantic space. An inverted index construction algorithm and the similarity definitions between vectors are proposed to calculate the similarity between two documents on the semantic level. The experimental results on benchmark corpus demonstrate that the proposed method promotes the evaluation metrics of F-measure.

In order to satisfy the needs of people's intelligent home environment, this paper proposes an intelligent home control system based on gesture recognition technology. To obtain and recognize gestures of human by the depth data, skeleton data and 3D point clouds uses Kinect. The Arduino microprocessor is used to process the received data to realize the intelligent control of home appliances. The body mass index BMI was generated by the acquired biological characteristics, and detects the user's physical condition. The experimental results show that the system can achieve effective control of household appliances and accurately measure human biological characteristics by receiving and recognizing human body posture. It proves that the system is innovative and practical.

With the development of image sensor technology, multi-sensor image fusion technology emerged and was widely used in the field of military surveillance, medical diagnosis, remote sensing, intelligent robot and so on. However, the current image fusion technology mainly focuses on the research of gray images, the color image fusion is rarely. Because color image contains more information compared with gray image, the research on color image fusion technology is becoming more and more urgent. In this paper, the realization of several typical color image fusion algorithms were discussed, the principle and their respective advantages and disadvantages were analyzed. Secondly, according to the different characteristics of visible image and infrared image, this paper proposes a color image fusion algorithm based on Curve let transform, this algorithm will combine visible image, infrared image with its negative respectively fusion, color mapping rules are in couple with the human visual characteristics. Experiments show that color fusion images obtained are richer in color, they contains more details and recognize easily

No abstract