An Efficient Image Depth Extraction Method Based on SVM

Compared with the two-dimensional media, the image depth of three-dimensional media can offer more intuitive and real scenes to the audience for feeling. But with the development of 3D display machines, there is a serious contradiction between the rapid of the 3D display machines and the lack of resources for the machines. To solve the problem, proposed an efficient image depth extraction method which utilizes the support vector machine (SVM). The label is established from the true depth of different videos, and the vector feature, haze is utilized as the feature vector. The training set is divided into a number of small training sets, reducing the sample size of the training sets, in order to improve training speed. The experimental results show that the algorithm is effective.