01-08-2014, 11:23 AM
PREDICTION OF RAINFALL AND CLOVD ESTION
INTRODUCTION
In this fast moving Digital Universe, the digital images play a major role in many applications. Our application here is the digital cloud images. One of the most interesting features of Earth, as seen from ground, is the ever-changing distribution of clouds. They are as natural as anything we encounter in our daily lives. As they float above us, we hardly give their presence a second thought. And yet, clouds have an enormous influence on Earth’s energy balance, climate, and weather. Our work is to recognize the type of cloud and to estimate rainfall. Rainfall plays an important role in the water cycle by providing water to the surface of the Earth. Rain sustains agriculture and provides water to streams, which is important for aquatic life and navigation. Excess rainfall, however, can be quite hazardous by causing flooding, which is a significant threat to both life and property. Because of the important role of rainfall in many aspects of life, it is not only worthwhile to observe where rainfall has occurred and how much has fallen but also to forecast rainfall.Even small changes in the abundance or location of clouds could change the climate more than the anticipated changes caused by greenhouse gases, human-produced aerosols, or other factors associated with global change. In order for scientists to create increasingly realistic computer simulations of Earth’s current and future climate, they’ll have to include more accurate representations of the behavior of clouds. The digital cloud images are collected and stored in file system. The sky status is found first and then the cloud status is found. With the cloud status we recognize the cloud type. The information about the type of cloud is given and the rainfall is estimated from the types and the color. The performance is calculated with various techniques.
SALIENT FEATURES OF ANDROID
Android is a mobile operating system that is based on a modified version of Linux. It was originally developed by a startup of the same name, Android, Inc. In 2005, as part of its strategy to enter the mobile space, Google purchased Android and took over its development work (as well as its development team). Google wanted Android to be open and free; hence, most of the Android code was released under the open-source Apache License, which means that anyone who wants to use Android can do so by downloading the full Android source code. Moreover, vendors (typically hardware manufacturers) can add their own proprietary extensions to Android and customize Android to differentiate their products from others. This simple development model makes Android very attractive and has thus piqued the interest of many vendors. This has been especially true for companies affected by the phenomenon of Apple’s iPhone, a hugely successful product that revolutionized the smartphone industry. Such companies include Motorola and Sony Ericsson, which for many years have been developing their own mobile operating systems. When the iPhone was launched, many of these manufacturers had to scramble to find new ways of revitalizing their products. These manufacturers see Android as a solution — they will continue to design their own hardware and use Android as the operating system that powers it. The main advantage of adopting Android is that it offers a unified approach to application development. Developers need only develop for Android, and their applications should be able to run on numerous different devices, as long as the devices are powered using Android.
PURPOSE AND SCOPE
The main purpose of this project is to recognize the type of cloud and estimate rainfall using certain features from the digital cloud images. Use of wavelet to separate the points needed for the cluster and using k-means clustering to combine those points provides a better performance than previous techniques used.
The Scope of RAINZAPP is predicting rainfall using the digital images rather than using the satellite images. The Satellite images costs lot so everyone can’t get them easily. When some new techniques are used we could get more accurate prediction
SURVEY ON CLOUDS
D.K. Richards and G.D. Sullivan describes the methods for using color and texture to discriminate cloud and sky in images captured using a ground based color camera. Neither method alone has proved sufficient to distinguish between different types of cloud, and between cloud and sky in general. Classification can be improved by combining the features using a Bayesian scheme. Malay K. Kundu and Priyank Bagrecha proposed the feature Extraction algorithm is a very important component of any retrieval scheme. The M- band Wavelet Transform based feature extraction algorithm is explained in this paper. Kuo-Lin Hsu, X. Gao, and Soroosh Sorooshian proposed some experiments. It shows that cold- topped cloud pixels with the same values of infrared brightness temperature may belong to different cloud type, thereby, indicating different rain rates at the underlying ground surfaces. It is suggested that the relationship between the satellite cloud-top brightness temperature and surface rainfall rate are non-unique for most pixel-based rainfall estimation algorithms. A scheme is developed, which first classifying cloud types based on the texture features of regional cloud images, then regressing the relationships of cloud brightness temperature and surface rain rate respective to different cloud types using the radar rainfall data. With the separation of cloud-texture types, estimated rainfall rates can be improved. A cloud-texture classification approach is introduced to process cloud images and estimates the surface rain rate underlying a cloud pixel referencing the cloud-texture type of the pixel. Instead of determining the surface rain rate based on cloud brightness temperature at a local pixel, as many rainfall estimation algorithms do (see Hsu et al., 1997; Bellerby et al., 2000), this approach extracts the features of cloud texture in a 4° × 4° window to classify the cloud imagery into a number of cloud (texture) groups. The relationship between rainfall rate and cloud pixel brightness temperatures at each assigned cloud-texture group is identified separately using ground-based radar rainfall data. Liu Jian and Xu Jianmin[4] describes an updated operational cloud detection method of FY-2C. Compared with FY2B three channels
EXISTING SYSTEM
The existing system is satellite system to predict rainfall using cloud images. With the advent of geostationary weather satellites in the 1960s and 70's, positioned above the equator at 5-6 positions around the globe to provide complete coverage, various techniques have been developed to estimate rainfall from visible and infrared (IR) radiation upwelling from the Earth into space. The higher the cloud albedo, the more droplets and/or ice crystals it contains and the deeper it tends to be, so the more likely rainfall is on the ground. And the lower the IR brightness temperature, the higher the cloud top, and the more likely the rainfall. A combination of both channels works best . Imagine a fair day with cirrus clouds, for instance. The IR channel may flag this as wet, because of the cold cloud tops, however cirrus is optically thin, so in the visible channel it is dry.The visible/IR rain retrieval algorithms work best at low latitudes, because at higher latitudes the view is more slanted, confusion arises with high-albedo surfaces of snow or ice, and deep-convective precipitation is less common. Another problem is incomplete pixel filling for small cumulonimbus clouds.At night no visible imagery is available. One can then use an empirical relationship between cloud-top temperature (deduced from the outgoing radiation in the 10.7-micron waveband), the simultaneous precipitation rate inferred from surface radar reflectivity, and the humidity profile (derived from radiosonde data). The rainfall rate (R mm/h) depends on the cloud-top temperature (T degrees Kelvin).
PROBLEM IDENTIFICATION
The purpose of this project is twofold: to recognize the type of clouds with certain methods and to estimate rainfall from certain observations. Observing rainfall is useful for evaluating moisture available for agriculture and determining how much will run off the surface into streams and rivers. For example, in moderate amounts, rainfall is necessary to grow crops, fill reservoirs, and maintain flow in rivers for navigation and shipping. However, in excess, rainfall runs off the surface in large amounts and can cause streams and rivers to overflow their banks and flood. Additionally, runoff can cause transport and loss of sediment and chemicals
SYSTEM DESIGN
System design is the process of developing specifications for a candidate system that meet the criteria established in the system analysis. The basic goal of system design is to plan a solution for the problem. This phase is composed of several systems. This phase focuses on the detailed implementation of the feasible system. It emphasis on translating design specifications to performance specification. System design has two phases of development logical and physical design. During logical design phase the analyst describes inputs (sources), outputs (destinations), data stores and procedures all in a format that meets the user requirements. The analyst also specifies the user needs and at a level that virtually determines the information flow into and out of the system and data resources. Here the logical design is done through data flow diagrams. The logical design is followed by physical design or coding. Physical design produces the working system by defining the design specifications, which tell the programmers exactly what the candidate system must do.
CODING AND TESTING
Software testing is the crucial element of the software quality assurance and represents the ultimate review of application, design, and coding. Testing represents an interesting anomaly for the software. During earlier definitions and development phases, it was attempted to build software from an abstract concept to tangible information. The testing phase is a very important phase since it is in this phase; we make sure that the system will perform the task without any error. Testing is vital to the success of the system and is being done by classifying it in to two ways-System testing and Program testing. Program testing involves checking the syntax and logic of the program. This checking resulted in achieving error free programs.
No matter how a programmer designs and plans application, the programs are sure to have a few bugs in them. Errors in the program immediately stop program execution and display an error message if the errors are syntax errors. After debugging one can identify the limitations of this project and hence corrections are made. During the system development, each source code was tested for its level of correctness. Each form was run a number of times in order to ensure that the details are entered correctly and works properly.
REMAINING AREAS OF CONCERN
The existing system is satellite system to predict rainfall using cloud images. With the advent of geostationary weather satellites in the 1960s and 70's, positioned above the equator at 5-6 positions around the globe to provide complete coverage, various techniques have been developed to estimate rainfall from visible and infrared (IR) radiation upwelling from the Earth into space. The higher the cloud albedo, the more droplets and/or ice crystals it contains and the deeper it tends to be, so the more likely rainfall is on the ground. And the lower the IR brightness temperature, the higher the cloud top, and the more likely the rainfall. A combination of both channels works best . Imagine a fair day with cirrus clouds, for instance. The IR channel may flag this as wet, because of the cold cloud tops, however cirrus is optically thin, so in the visible channel it is dry.The visible/IR rain retrieval algorithms work best at low latitudes, because at higher latitudes the view is more slanted, confusion arises with high-albedo surfaces of snow or ice, and deep-convective precipitation is less common. Another problem is incomplete pixel filling for small cumulonimbus clouds.At night no visible imagery is available. One can then use an empirical relationship between cloud-top temperature (deduced from the outgoing radiation in the 10.7-micron waveband), the simultaneous precipitation rate inferred from surface radar reflectivity, and the humidity
CONCLUSION
The type of cloud is recognized and the rainfall is estimated by using the novel methods like Wavelet and K-means Clustering. The performance is better compared to the previous techniques like Laws and other clustering techniques. Considering the cost factors and security issues, the digital cloud images were used to predict rainfall rather than satellite images. The status of sky is found using wavelet. The status of cloud is found using the Cloud Mask Algorithm and histogram equalization. The type of cloud can be evolved using the K- Means Clustering technique. The type of rainfall cloud is predicted by analyzing the color and density of the cloud images. The cloud images are stored as JPEG file in the file system. Analysis was done over several images. The result predicts the type of cloud with its information like classification, appearance and altitude and will provide the status of the rainfall. In future the accuracy can be increased by using other transforms like curvelet, contourlet etc. The parameters like dew point, temperature, wind direction, humidity and precipitation can be included to in crease the performance. Certain specific rainfall estimation algorithms can be used for getting the result in a dynamic way.