Gesture detection by virtual surface
1 like108 views
1) The document proposes a method for gesture detection using a virtual surface detected by a webcam or front-facing camera on a laptop or smartphone. 2) By tracking the number and position of pixels representing an object's shape at different distances from the camera, the computer can detect movements of the object maintaining a constant distance as gestures on a virtual surface or plane. 3) This technique aims to enable gesture control of computer functions like mouse movement or app launching without requiring specialized cameras, as a cheaper and more portable alternative to physical input devices.
Gesture detection by virtual surface
- 1. SYNOPSIS ON Gesture detection by virtual surface Submitted by :- Ashish garg, CS Harsh, CS Vishal Singh, CS Shubham Goel, CS
- 2. GESTURE DETECTION BY VIRTUAL SURFACE Abstract : Our eyes gives us great sensing power to interact with the world around us. From the five senses, it is most important one as it become very difficult for anyone to live without them. We are able to understand the motions that takes place around us easily by simply see them. Likewise the researchers have been working on gesture detction for quiet a long time such that it become possible for the computer to understand the operations that we perform through keyboard and mouse by just moving head or hands. Also it gives more features to its user and a convenient way to use anytime anywhere. The requirement of external devices diminishes and also the space acquired by them. Camera is an eye for the computer to see around to scan the real world and map the information to an image. It is a sensor to gather the data of real world in the form of images. And this information or images are valuable as it can be perform different tasks on the machine. Objective : Our objective of this project is to map the user actions with that of a mouse using the webcam of laptops or front cam of smartphones. As we all know that already there have been so much development in areas of computer vision, but all these things have become possible by using powerful devices such as depth sensing camera and infrared sensors. The central theme of the project is bring down the requirements of external devices for providing the computer vision.
- 3. Idea : 1. Undersatnding a Natural Phenomenon In school days, we are all taught that a object of definite size appears to be of different sizes depending on the distance of the object from the viewer and the viewing angle. Suppose a plate with flat surface is observed by a person keeping the plate at different distances. When the surface was far away from the person, it appeared very small and as the distance between the person and plate surface is decreased, the surface begins to appear big. 2. Natural Phenomenon being followed by a camera A camera scans the space in front of it and saves the data in an image in the form of pixels where each pixel represent some part of the outside world. Now think that the same flat plate is kept in front of the camera at different distances. If the plate is 20 metres away, camera captures an image in which a small number of pixels represent the flat plate. But if we decrease the distance between them and captures an image, then the number of pixels representing the same gets increased. On further shortening the distance, number of pixels will increase proportionally. From this, it can be concluded that the number of pixels required by an object in an image is directly proportional to the distance between the object and camera. P ( number of pixels) d ( distance between camera and object) 3. Detecting the object From the above result, we found that an object acquires a specific area of pixels in an image for a particular distance from the camera. Also when the distance changes, the number of acquired pixels also gets changed. But the shape of object representation never changes in the whole scenario.
- 4. An important point to be noted is that the object representation area extends or contracts from its boundaries whenever the distance is changed, that is, the changes occurs only on the pixels that are near to the boundaries of object representation and neither change occur at the core pixels nor the position of core pixels of object representation is changed. This is illlustrated in the diagrams. The shape of object should remain same for facilitating the object detection. In order to detect the object, its shape should be defined to the system such that it creates two frames of same shape. Once the frames are created, these can be used for object detection by applying the following two rules : 1) the pixels representing the object must cover the inner frame area. 2) the pixels representing the object must not exceed the outer frame area. For the object to be detected by the machine, the object representing pixels must satisfy both the above conditions.
- 5. 4. Implementing the algorithm The object is at a particular distance from the camera and also the area covered by the object representation in the image satisfies the two conditions mentioned in previous section. So, the computer is sure about the presence of object. Now if we consider that there is a virtual plane in front of camera at the same distance as the object, that is, the object and virtual plane coincide. Whenever any movement are made by the object horizonally or vertically but maintaining the same distance from the camera, the computer can detect those movements as the bunch of pixels representing the object in the image are re-positioned together every time in the same direction and this helps the computer to detect the gestures made by object on the virtual surface or plane.
- 6. 5. Mapping of motions As the motion of objects can be detected by the camera, various motions can be defined in the computer system which will be used as a signal to perform a specific operation and the system become able for executing the tasks just by moving fingers or hands in front of computer. Most importantly, the hardware devices are replaced by a piece of software which reducing the cost of computer and make it more portable. Scope : After a few years, the devices like mouse, keyboard and other physical devices will not be an elligent or smart way to use computers as they require space and also counts for weights. For using these devices, it requires a proper setup.Whereas the virtual technologies are getting more popular due to ease of use and high extensibility. The virtual environments are created around the computers which facilitate the user to perform the tasks in more sensible and convenient way. The external devices will be replaced by small software and promotes the resource utilization of sensing devices like camera. The industry benefits the users by provide advance features at a low cost. Conclusion : The concept of using the surface area of an object as a tool for gesture detection reduces the requirement of depth sensing camera to perform simple task such as mouse operations, changing volume up or down, starting a specific service or software by making a specific registered gesture etc. It provides the machine, the ability to interpret the user's actions and perform the specified task. This allows the programmers to add extra functionalities of computer vision to the computer through softwares.