The chapter on avoiding piracy will be revised today and tomorrow

The significance of the target positioning of the S screen is in the process of information development to create all kinds of software systems, such as customer management system, financial management system, warehousing system, personnel management system, information management system, etc., often developed by a number of software developers, but the software system often needs to exchange data and information, and some business processes must be processed by multiple systems to complete. Pen, fun, pavilion www. biquge。 info However, users generally do not grasp the relevant development data of these system software, and it is difficult to customize the data exchange system according to the needs, resulting in the circulation between data needs to be operated manually, and it is difficult to meet the needs of office under the condition of informatization. Reverse engineering analysis of related systems can obtain the underlying data structure of the system to develop the corresponding data exchange system, but the versatility is poor, and the cost of reverse engineering analysis is also expanded with the continuous development of anti-tracking technology. Machine vision converts the target into digital information of the image through the image acquisition equipment, and is equipped with the corresponding algorithm module to extract multiple feature points from these digital information and analyze and judge, simulate the manual data exchange operation, and improve work efficiency. The object recognition system based on image feature information extraction has been widely used in various fields, and this paper provides a human-computer interaction module for software robots by establishing a simulated manual operation architecture for extracting and judging the target object information on the Windows screen. 2. Feature matching of screen objectsScreen objects refer to the various keys, windows, pictures, icons, mice, etc. displayed in the human-computer interface of various applications in the Windows system. These objects are derived from object-oriented code, have code features, are ultimately used for on-screen displays, and have geometric features. The target object positioning of the Windows screen is mainly based on the geometric features of the object, which requires the use of image matching algorithms. There are many kinds of image recognition matching algorithms [1], which have high requirements for their reliability, accuracy, and real-time performance. In recent years, relevant scholars at home and abroad have developed many practical algorithms, which have led to the rapid development of this technology and its successful application in various fields, such as object tracking, fingerprint recognition, note identification, image stitching, etc. Image matching technology can be divided into feature-based matching, grayscale-based matching and other methods according to the feature level. The grayscale feature matching principle is mainly based on the grayscale information of the image content to measure the degree of similarity, which is mainly used in some specific occasions. The feature matching method needs to first segment the image into points, lines, and surfaces, and then extract the features, and match the similarity of the extracted features and establish a mapping relationship. The architecture of this paper is mainly based on the SIFT (Scale Invariant Feature Transformation) algorithm based on feature extraction[2]. Figure 1 shows the flow of the image matching algorithm. Figure 1 Image Matching Algorithm Flow The SIFT algorithm was proposed by David Lowe in 1999 and further refined in 2004. It can handle the scaling, rotation, affine transformation of the graph and keep it well matched. In the comparative study of related local feature description algorithms, it is confirmed that SIFT and related improved algorithms have considerable robustness. The SIFT algorithm detects the local features of the image, which has the characteristics of uniqueness, multiplicity, real-time and strong scalability. The essence of the SIFT algorithm is to locate the feature points on different scales obtained by using Gaussian convolution, and determine their feature descriptions and match them based on the local gradient direction of the image. 3. Application scheme 3.1 grayscale first needs to take a screenshot of the entire screen to be extracted, and the image obtained from the screenshot is stored in BMP (bitmap) format. The screenshot image has three colors: R, G, and B, and the numerical range is 0 to 255, in order to facilitate processing, the image needs to be processed into a 256-level grayscale image, and a common method is used here to convert, and the formula is as follows: grayscale value = 0.3B + 0.59G + 0.11R3.2 SIFT positioning uses the SIFT algorithm to locate each element of the Windows screen program, and the characteristics and attributes of each target object need to be obtained in advance. Matching is then made based on the model that has already been built. In the process of feature selection, due to the excellent characteristics of the SIFT algorithm, there is no need to consider too many influencing factors such as translation, rotation, scale, brightness, etc. The steps of the SIFT algorithm [3] are divided into five steps: scale space construction, extreme point detection, extreme point positioning, feature point direction allocation, and feature point generation. First of all, it is necessary to establish a target feature database, and extract a large number of feature points such as buttons and windows in the application interface through analysis and add them to the target feature database for feature matching. Due to the differences in the geometry of the buttons, windows and other elements of each application, there may be great differences in the feature points, and the target feature database also needs to have a learning function to continuously collect new feature points and add them to the feature database to improve the accuracy and stability of recognition. 3.3 Text positioning[4] After obtaining the application interface element object, it is also necessary to determine the position of its Chinese text and identify it, and the current common text positioning algorithms have roughly three categories: one is the edge-based detection algorithm, which uses text edge information and local histogram positioning; The second is a texture-based detection algorithm, the core idea of which is to treat text as a special kind of texture segmentation detection, which is suitable for text localization in complex background situations but has low efficiency. The third is based on the extraction algorithm together with the region, which uses the regional geometric conditions to set the threshold to extract the range of the text area, and cannot accurately extract the text communication area under complex background conditions, and the application surface is relatively narrow. Combined with the characteristics of the texture detection algorithm and the connected area method, this paper first roughly obtains the approximate connected area, and then obtains the texture features through the wavelet algorithm. However, from the author's understanding, many data security management personnel are not proficient in the proper security management skills, and it is difficult to make effective use of information resources, precisely because of the lack of professional operation ability, so that data security problems within the social security system occur frequently. 2. Failure to implement security measuresAt present, the dynamic and comprehensive characteristics of modernization are a new requirement for data networks, and it is precisely because of the existence of the above requirements that hidden dangers are buried for data security. Without fully understanding the rules of network use, the insured people directly enter the operation, so that some sensitive data is exposed, and there will be a risk of intrusion into the system. 3. The comprehensive lack of data network of the solution is constantly changing, but in most of the social security system there is no change in the network and strengthen their own security management, which will strengthen the security risk to a certain extent, and the formulation of relevant solutions is also difficult to show the comprehensiveness, which has caused a certain false sense of security in the social security system, for a long time, it will inevitably reduce the vigilance of relevant managers. In fact, the data information within the social security system is constantly changing and increasing, if it is a long-term use of a maintenance management program, this is unscientific, it needs to be understood that firewalls or anti-virus software can not completely solve the network security problem, even if a large number of use of security protection products can not completely eliminate the hidden security problems. At this stage, some data security maintenance companies in foreign countries have fully realized this, and they have added their own solutions to the social security system of their own countries, in short, the role of anti-virus software providers has become the role of security program formulators, obviously under the guidance of the professional maintenance team, the problem of data security will be gradually solved, and China should also learn from it. 4. Lack of prevention mechanismFor the social security system, the relevant person in charge has not established the data security system from the perspective of the security prevention mechanism, which is naturally difficult to play a deep role in the security guarantee of data information. For the operation of the entire social security system, if the corresponding protection system and security inspection are not formulated, it means that data security problems may occur at any time. 2. Analysis of solution measures 1. Data security management work is strengthened for us, the social security system is not unfamiliar, whether it is unemployment insurance, pension insurance, or medical insurance, etc. are included in the social insurance system, as the staff of the social security system, they are the protectors of the personal accounts of the insured, so the first thing to do is to strengthen the relevant data security management, from the root. In order to achieve the goal of security system management, it is necessary to carry out security assurance at multiple levels. The author believes that the following measures can be taken to strengthen data security: in the process of transmitting data information, it is necessary to protect the data boundary by means of account real-name system, ID card identification, etc., and then protect the information in the data layer and the link layer by means of access and filtering, and can be appropriately used to serve the proxy service in the transport layer, so that the insured can verify it through digital signatures or data encryption and other forms when operating, so as to protect the security of data. 2. The establishment and improvement of laws and regulationsAt present, the domestic law on the data security protection of the social security system still needs to be improved, and it is still difficult to implement and general, so the government functional departments should establish effective laws and regulations under the understanding of the actual data security maintenance, and continuously improve in practice to ensure the social security system