We are researching on a measure of picture quality considering perceptual properties. We have defined a measure of picture In this Paper we quality using a linear combination of 6 factors . studied the dependency of a factor to the other and contribution of factors to measure of picture quality. We have discussed in detail how to j catch the loal feature of image. Consequently , we have classified following three groupes of factors. l) Random noise 2) Block distortion 3) Distortion at contour
It has been for a long time since S/N ratio isn't always good at picture quality i evaluation. However, we still often evaluate picture quality by S/N ratio. Because, it is easy to calculate and it has a relation to picture quality. Furthermore good alternative isn't established yet. In this report, relationship between picture quality and Luminance Differential Information (LDI) is described. As a result, it turned out that distribution of LDI, which depends on the coding met,hod, correspond to the appearance of the distiortion. And it is useful for picture quality evaluation.
It Is necessary to reduce resolutions of pictures for transmitting video signal at low bit rate. Therefore, the subjective quality of pictures with impairments due to the coarse resolutions of space, time and gradation is investigated. The following results are obtained; i)as viewing distance increases, the spatial resolution degradation becomes less annoying. but the temporal one becomes rather annoying, ii)the formula for summation of multiple impairments proposed by Lewis et al. applies to these degradations approximately.
This paper discusses the coding errors of image coding in the uniform perceptual space, and discuss also the objective quality estimation considering the Human Visual System. The advantages of the coding in the uniform, Perceptual space are as follows; 1) If the image signal is proportional to the visual sensation, the coding error will be independent to the amplitude of input signal and ; will be proportional to perceptual disturbance. 2) If the objective quality measure is obtained based on the error described above , the optimal coding parameters will be calculated by the repeating computation. 3) An image coding can be designed systelltatically.
A new Y/C separator are proposed. In conventional television receiver's . Y and C are separated by a fixed BPF or comb filter ;i.e, the separation is a based on the statistical characteristic of the composite television signal and causes the cross-color and cross-luminance disturbances-Since the important information of image is on the local parts of pictures , the Y/C separation should be accomplished by an adaptive method which extracts local features in the spatial domain. An adaptive Y/C separator is designed and made completely by digital hardwares. Good results are obtained.