Text Localization Algorithm And Its Application In Intelligent Reading Software

With the continuous improvement of people's living standards, as well as the development of multimedia technology, mobile terminal products with camera function has been widely spread, which is the most representative of the smart phone, which makes it easier to capture images in everyday life. Mobile phone performance continues to increase, so that through the smart phone can handle a lot of the past must be done through the personal computer and the scanner to complete the work.In addition, the aging population has become a prominent social problem in today's world. Compared to the developed countries in Europe and America, the aging problem is more serious in China. With the increasing of the aging population the elderly's spiritual and cultural life can not be ignored anymore. As human nutrition, books are helpful to enrich the spiritual life of old people, but because of vision recession, limited energy, there are reading obstacles for the elderly.1.based on the development of mobile phone cameraUsing the underlying camera interface and Surface View, the Android mobile phone camera is developed, the camera is embedded into the intelligent reading software, and the real-time acquisition of paper books document image is realized. And through a custom control to solve the problem of the size of the image acquisition, the size of the books and cell phone support is not matched on the length and width ratio.2. text location algorithmAiming at the problem of text location for static document images, two kinds of text location algorithm based on projection method and connected domain method are designed in MATLAB environment. In the Android terminal, take full account of the user experience, realize the dynamic text positioning effect under the preview state. And the acceleration sensor is used to monitor the mobile phone and ensure the validity of the collected images.3.text recognition and speech synthesisOn the basis of text location, the Fork-Join framework is applied to the smart phone, which is based on the location of the text line. The whole page of the image is segmented, and the speed of text recognition is improved obviously, and the recognition rate is improved to a certain extent. Software with Baidu voice to complete the speech output.