A A A Robert Koprowski - Kanał RSS

Image processing and analysis

Eyesight is one of the most important human sense apparatus. By means of eyesight we obtain from 80% to 90% information about surrounded environment. There is an analogy between image received by human eye and computer. There is a mosaic consists of many neurons in the back wall of the eyeball. These cells react to light and they send obtained information to the brain. Image in the computer monitor also consist of many points, called pixels. The colour and intensity of individual image points is a reflection of the image obtained from the optoelectronic transmitter (…).Our sense of sight and the way of its interpretation is certainly more sophisticated than the computer image interpretation. Based on the earlier work [12] the following disciplines where “artificial vision” is necessary can be defined, These disciplines are:

 

  • automatics (improvement of the robots autonomy that they can replace people in many mechanical work; quality improvement of the systems of automatic control),

  • medicine (automatic analysis and recognition of tissue samples images, roentgen images, ultrasound images and images obtained by using tomography and magnetic resonance, etc.)

  • forensic (identifying the image attributes witch are undistinguishable for human eye),

  • geodesy and cartography (automatic processing of huge image data, especially aerial and satellite images),

  • transportation (detection of traffic direction and intensity, defining the size of car queue on a intersection, automatic detection of collision and accidents),

  • research laboratories (control of materials, product quality control),

  • Security of objects remaining under special supervision (luggage control, traffic surveillance in an area, detection of intrusions)

  • Public defence/military (directing of intelligence weapons and its navigation on a modern battlefield),

  • astronomy and astrophysics (analysis of spectral images overstepped the human perception),

  • metrology (measurements made based on the image),

By analyzing the use of the image processing in above disciplines it can be concluded that despite many imperfections computer based image processing has many advantages over human abilities. The principal benefits of using computer based image processing are [10]:

 

  • lower price of image analysis,

  • result reproducibility,

  • the ability to record the results automatically,

  • varieties of programming and processing methods depending of user needs

  • lack of human factors such as fatigue, boredom, stress, bribery, etc.

  • work in any spectrum of light (visible, infrared and ultraviolet)

  • work in places inaccessible to human

  • easy and inexpensive duplication of tested solutions,

  • lack of instincts and self-preservation (weapon),

  • Matlab as a development environment for image processing

    MATLAB is used to carry out engineering and scientific calculations. It allows for the modelling, analysis and visualization of data and calculation results. Commercial success of MATLAB is based on its flexibility and ease of expansion by the user.

  • Image aquisition

    Image acquisition is the method of converting visual information about the world around us into digital form, convenient for processing in computer systems.

  • Image in Matlab

    One of the main features of the image is the range of colours. One can distinguish black and white, monochrome images (called multi-shade) and colour images.

  • Discrete image structure

    The fundamental issue that occurs in image processing, when image is obtained from the CCD, is to define the parameters of the artificial representation of the image.

  • Point Transformations

    The transformation point, also known as transition non-context, are the transformation of gray level or colour saturation of each image point separately.

  • Regional Processing

    During the image processing often it is important to apply the changes to selected fragments (regions) of the image.