Image Recognition Method Based On Knowledge Element

With the rapid development of digital technology, the amount of image data into geometric growth, and manual identification of image data has been difficult to meet to keep up with the pace of development of image acquisition which makes the image recognition requires a qualitative development. Vision in the field of image recognition is the ability to talk to the human visual system to achieve comparable recognition efficiency, but in the image recognition process, purely from the perspective of the image on the image content recognition has been difficult to meet the demand, simply increasing the sample label and increasing the amount of computation has been difficult to make accurate identification of a qualitative improvement; and because the human eye itself is an imperfect image recognition system, which is really well-developed brain reasoning recognition system, so the knowledge of image recognition reasoning become a breakthrough image recognition. But the image of knowledge representation and reasoning is far from perfect, is still a hot research technology to improve the image recognition.For lack of practice the above problems and present research paper from the perspective of the knowledge element theory, the first knowledge of image recognition to classify, to distinguish the underlying image and high-level semantics. Depending on the properties and then build a knowledge element model of knowledge of image recognition and experimentally verified.In this paper, the image recognition field has certain theoretical significance, using knowledge element of the image knowledge systems knowledge representation, knowledge can not only take advantage of many features of dollars into the possible complete knowledge of image expression, and easy on the existing knowledge reasoning; because knowledge element has its own complete theoretical system, so take advantage of the knowledge element of the image of the knowledge system knowledge reasoning, is easy to find with its internal implied association; knowledge element has been used in the literature management, mechanical design and emergency management other fields. If the knowledge element can be applied to image recognition, that is the extension of the application of the knowledge element.There is also some practical significance on the practice field of image recognitions, image recognition has been widely used in various industries such as medical, security and so on. Image recognition using knowledge inference can further enhance its recognition accuracy, especially as the field of security. Because there are massive security image data identified under artificial conditions, a huge human cost, and the proportion accounted for critical data in a very low amount of data in the data, the human cost of waste. Image by image recognition system identification methods, learning systems also require a huge amount of labeled data, but also in the recognition process complex scenes also unsatisfactory, so the image is necessary to identify the knowledge-based reasoning practical significance.