In this paper, It created the concept of smart doorbell. The new product is not in the existing market, The IT market was extended to the popularity of smart-phones such as Internet of Things(IoT), home automation in age of one hand computer. When gue...
In this paper, It created the concept of smart doorbell. The new product is not in the existing market, The IT market was extended to the popularity of smart-phones such as Internet of Things(IoT), home automation in age of one hand computer. When guests visit the house is empty, visitors to notify the landlord that the smart-phone push notifications, and in real time check the face of a visitor through the application, It is possible to open the door remotely, regardless of the distance. Also it introduced a system that can open and close the door, looking for a feature point of the face taken to compare with existing data through the camera installed in the doorbell by using a kind of image processing SIFT.
When using the Wi-fi module instead of the Ethernet module escape from the limitations of a wired. By designing the embedded device and a communication system with low power consumption is expected to be smaller as the conventional door lock. Push-on is used a public server Google. But if you are building a server, it seems to be for a system based on the application to be able to take advantage of the push notification and video support.
In the case of keypoints extraction through the SURF, the speed was improved algorithms sift. But this delay time is to make the matching results and existing data. And the lower the number of keypoints to be matched when using real-time imaging. Thus, shoot an image at a constant illumination and make the matching Algorithms. It can be obtained more matching points. Use the probability to obtain the reliability of the determination of the data has not been quantified. After, Separate the background and character of the recorded image, and if the process of the feature points based only on the character created, it is possible to build this matching system than perfect.