Human and Computer Vision
Human eyesight is a sophisticated human mechanism. The sophistication stems from the fact that over half of the human brain is dedicated to vision. Understanding the basic operations of the brain requires knowledge of the visual system. Computer vision is developed in the same way that the biological vision system of humans is. There are, nevertheless, major distinctions between the two vision functions. The discussion in this paper concentrates on the parallels and contrasts between the two visual systems. Furthermore, this article highlights the critical aspects and features of the human vision system that can be modified to improve the effectiveness of the computer vision process. Lastly, this discussion will highlight the aspects of human vision that should be ignored by computer scientists when building machine vision.
Similar Perception Process
Both human and computer visions have a similar perception process. When an individual reads a magazine and observes a picture, the human brain analyses the picture or the image in a clear and smooth pattern. However, close analysis of the picture indicates several minute dots of ink. The human brain combines the dots to develop a visual appearance of a smooth image (Umbaugh, 2016). The human brain is also competent enough to combine dots that have different colors to develop new images. This capability enables the newspaper outlets to utilize crude printing methods to generate pictures that can be easily recognized by the readers. A similar vision process is experienced in computers and robots. The pictures illustrated on the computer screen are shown as data or information pixels (Rautaray & Agrawal, 2015). In the same way the tiny dots of ink are crucial to the development of magazine pictures, so are the pixels in coming up with computer images. Additionally, the brain can analyze a scene that has several unique objects. This is done through identifying and focusing on the most significant part of the scene. The focus on the key areas of the scene is important because the brain is not capable of analyzing all the objects at once as it can not receive detailed information from everything. The brain develops a visual image of the scene to locate the areas that are visually important or interesting. The computer system is also selective when analyzing an image. The entire image is entered in the computer as a data or information input. However, the computer focuses on the most significant information that is relevant to a given scenario and subjects it to analysis. As such, it is able to ignore clutter that is not necessary in the analysis. Therefore, both human and computer vision systems are similar because they target the important parts of an object or image for analytical purposes (Granlund & Knutsson, 2013).
Superiority of Human Vision System
In today's world, computer vision that currently exists is less superior to the human vision system. The complexity of the human brain is related to its capacity to analyze the world in more than 30 unique models with the ability to process information in different patterns. The patterns of processing the data include texture, color, smell, or shape. How the human brain perceives the world is a result of the reverberations of the visual signals between the different categories of the sensory system. The existence of the visual illusions contribute to the sophistication of the human visual system. The difference between the human and the computer visual perceptions is referred to as the semantic gap (Granlund & Knutsson, 2013). It is challenging for a computer to achieve the same level of effectiveness as the human vision system. The explanation for this reduced efficacy of the computer system is the reliance on the programming that is made use of by the computers. The vision system of the computer relies on the mathematical models of programming that have been designed to function in a specified manner. In other words, for the computer to analyze the shape and patterns of the pixels that are significant in the development of computer information and images, there must be a utilization of mathematical formulas.
Adopting Human Vision Qualities
There are several aspects of the human vision system that computer scientists can adopt to improve the functionality or the capabilities of the computer vision system. The most outstanding quality of the human brain is its capacity to use only a small amount of data and energy to efficiently process information. While computer scientists have developed computers with the ability to process information faster and store more data than the human brain, they have not managed to rival the brain in terms of efficiency. Computers rely on the processing abilities determined by the algorithms which are programmed by humans. The algorithms should be developed to analyze visual information within microseconds. To improve vision system of the computers, the scientists should be seeking on the ways to enhance the efficiency of the systems to ensure use of less energy and the amount of data required to analyze information (Sonka, Hlavac & Boyle, 2014).
Ignoring Negative Aspects of Human Vision System
Scientists should ignore the negative or ineffective aspects of the human brain and vision system when developing the new computer and robotic vision systems. One key problem with the human brain is that it is affected by certain errors when processing information (Beck, Hope & Rosenfeld, 2014). The human brain is effective when processing information about common situations. The concept of optical illusion where the brain portrays an incorrect version of images or information as being real. Some of the examples of optical illusion include angle distortions, contrast of movement, interpretation of 3D among others. In trying to build super computers, the scientists have managed to eliminate these types of errors. However, they may occur if the mathematical models employed in developing the programming are faulty.
References
Beck, J., Hope, B., & Rosenfeld, A. (Eds.). (2014). Human and machine vision (Vol. 8). Academic Press.
Granlund, G. H., & Knutsson, H. (2013). Signal processing for computer vision. Springer Science & Business Media.
Rautaray, S. S., & Agrawal, A. (2015). Vision based hand gesture recognition for human computer interaction: a survey. Artificial Intelligence Review, 43(1), 1-54.
Sonka, M., Hlavac, V., & Boyle, R. (2014). Image processing, analysis, and machine vision. Cengage Learning.
Umbaugh, S. E. (2016). Digital image processing and analysis: human and computer vision applications with CVIPtools. CRC press.
