And so the exhibition had done successfully and we are really grateful for your visit! Also its always nice to see student/public that are interested in our project and asked us a lot of technical questions. During the 3 days exhibition, we really had a great experience as we received a lot of suggestion on how to continue to improve our DIY microscope, the way to market it and the best feeling is by knowing that we are the inspiration for school teacher/parent on what to let their student do for their school science project. Again, we, team sotong really grateful and appreciate for all of your support and interest!
Here's some photo of the exhibition day ! XD
So now, the question is that, will we just stop here???
The answer is a big NO!
Following from the project report and internal presentation done here in our faculty, we are told to further improvise on our microscope! yeayyyy!! The mission given by our lecturer Dr. Woon is that to make the microscope lever into auto focus mode. This meaning that, the up and down motion to bring the specimen into the focal length of the camera will now on it's way to become auto mode. Sounds interesting right?? For this part of the project, it will be conducted by our member, Chiam.
Therefore, week 9 start with some data collection on how to transform the lever into auto focus mode. In auto focus mode, there need to be "something" to tell the lever when to stop its up and down motion once the image is sharp and focused on the smartphone. Therefore, it is strongly believed that by doing some modification on the Arduino sketch (most probable by using boolean), we can "teach" the Arduino when to continue to adjust the specimen up and down accordingly and when should it stop by itself! Also, we need "something" to tell the lever what and when is the image considered focused and sharp enough so that there will be a standard. Once the mentioned standard is reached or achieved, everything should stop. For this, "image analysis" should be brought out to perform its action.
Image analysis is to process an image into its fundamental components in order to extract data. This is such that a kid trying to dismantle on a toy building built by Lego blocks of different sizes, shapes and colors, and then try to study the appearance on every single Lego block used to build the toy building. The tasks that are involved in image analysis is such as finding shapes, detecting edges, removing noise, counting objects and also measuring region and image properties of an object. A very apt example of image analysis is such as "Face Recognition" system that are available on the Iphone X launched recently. Besides that, the camera of our smartphone too uses the technique of image analysis, this can be seen when your camera focus on an image by itself or when you tapped it. Image analysis is getting wider and wider in the world and also in social media, as image is getting preferred to be used as communication medium nowadays.
Image Analysis
There are two types of auto focus(AF) available in our technology. The active AF and passive AF. Active AF uses an external sensor to tell whether the image is focused already or not. It's advantage is that it can even focus under situation with different lighting condition, however, due to the infrared or ultrasound emitted by the sensor, one may find it's not easy to focus on specimen with window or glass. Another type of AF is the passive AF. Passive AF relies only on image information to determine if the image is in focus. Therefore, worrying that our specimen might involve glass piece (for example, the specimen is sandwiched in between two glass pieces), the choice of using active AF is eliminated and passive AF comes to our priority.
In the initial phase of making the AF microscope, there is two rough plan in which can be taken into consideration. First is that to move the Arduino to be connected to smartphone directly and thus from there, controlling the lever up and down motion, and the second is that, to connect the smartphone to the computer in order to analyse the image in the smartphone (with the help of processing software, such as, MATLAB), and then, from the computer, the instruction and result is sent to the Arduino. Both of the method requires further research to determine which are the more convenient and effective method in directing and instructing the Arduino.
Auto focus are related to the sharpness function. Sharpness function is a quantitative description of the sharpness of the image in the mathematical term. It is a numerical measure that provides a value where the value stands for the degree for an out-of-focus image. As contrast of an image is directly related to the sharpness, therefore it is convenient for us to take contrast measure to solve for the sharpness function. Therefore, by using this technique, we believe that an image can be "set to a standard" where it is said to be "focused" and in turn to instruct the Arduino to continue adjust it's position automatically until this "standard" is obtained.
Different lens position gives different contrast. The optimum contrast is to be obtained when the lens position is in the focal length area. But then, for our case, the matter is not the lens position, but the specimen position as the lens position is fixed.
Xu, X., YingLin, W., JinShan, T., XiaoLong, Z., Xiaoming, Liu. (2011). Robust Automatic Focus Algorithm for Low Contrast Images Using a New Contrast Measure. Sensors, 11(1), 8282-8284. doi:10.3390/s110908281
MATLAB is a software used mainly for math, graphics and programming, where it suits all our needs. It is able to analyze data, develop algorithms and creating models. It is also able to signal processing, undergo computer vision and robotics to serve most of our needs.
That's all for the week 9 update. Stay tuned for more on the auto focus progress ya! See ya!
No comments:
Post a Comment