Currently,along with the rapid development of search engine and it towards the direction of the commercialization step by step, special search engine technology arises at the historic moment.In this article, according to the principlemodel and indexers of search engine tool of Lucene, we design a search engine system. In this system, a user can input a text in front of home page of this website, the system will use Ajax technology to match the search content automatically.And then it will show search results in the form list below the drop-down. At the same time we can click search, it will highlight the search content and display in jsp page. Actually,it has absoultely meaning in realistic.

Orthogonal Frequency Division Multiplexing(OFDM) is going to become the most important technique in the future communication system. The paper presents the principle of OFDM,establishes a simulation model of OFDM communication system based on simulink.By utilizing the system,we briefly analyse the simulation results and conclude the the performance characteristics of OFDM system.

The main content of this paper is based on the estimate of LWH(length width and height) dimensions of a machine vision moving vehicle. The information of a vehicle’s length width and height is an important parameter to monitor Road. To achieve non-contact rapid detection of the parameter of a moving vehicle’s size, can effectively improve the efficiency of the road monitoring and provide effective technical means for China's road transport management department. First get the vehicle in sequences image and extract the vehicle from background. Then, we transform the two-dimensional image to three-dimensional image with a simple projection model for getting the parameters of the length, width and height of the vehicle. Experiments show that the accuracy of the test results of this algorithm is relatively high.

Nowadays,artificial recognition is widely used in the mushroom inspection system, however, it depends on subjective judgment of inspectors.Therefore,the testing personnel's experience, technology and other factors will affect the objectivity and accuracy of test results.Commodity inspection system need a high-speed, objective and accurate method for the on-line hair detection in the mushroom.On the basis of summary of domestic and foreign research, this paper studies the target identification and feature extraction techniques based on computer vision, conducts a feasibility study for the real-time hair detection system.

By analysing the Mushroom image, the paper puts forward a kind of line-structure extraction algorithm combination of local gray value and continuity of line direction .After the operations in many aspects of basis image processing, such as gray-scale, denoising , segmentation, contour detection and morphological, this article has developed a set of hair detection system based on computer vision for the Mushroom. The experimental results show this system could well meet the actual needs, and has a broad market prospect.