Color machine vision

Filtering the information within a scene begins with matching the vision system to its industrial requirements. Just as humans can adjust to a variety of situations by dilating their pupils or by tuning themselves to look for a particular shape or color, machine vision systems must also be somewhat flexible. Typically, however, the most efficient system is one which is designed with only limited applications in mind. For this reason, machine vision designers have developed a variety of application-specific techniques and systems to meet the speed and accuracy standards that modem industry demands. 

Legeard, D., Marty-Mahe, P., Camilleiapp, J., Marched, P, and Leredde, C. (1999), Reed-Time Quedity Evaluation of Pork Hams by Color Machine Vision, in Proceedings of SPIE, Vol. 3652, pp. 138-149. 

O. Tokusoglu, M. Balaban, Correlation of odor and color profiles of oysters with electronic nose and color machine vision. J. Shellfish Res. 23, 143-148 (2004)... 

Correlation Between Color Machine Vision and Colorimeter for Food Applications... 

The goal of this research was to develop a rapid color monitoring technology for off-line, on-line processing of value-added food products. Color machine vision technology was used as a fast non-contact, non-destructive method for quantifying and predicting the color quality of processed foods. Specific objectives were to ... 

Accurate machine vision color measurement is a non-trivial task and every step in a color quantification process can affect the accuracy of the measurement. Major areas of the color machine vision system (CMVS) development task consists of acquisition of color images, low-level image processing to compensate for aberrations caused by qitical impurity and noisy electronic devices in a system, color quality index selection/develqiment, and defming relationships between instrumental and sensory evaluation results. Successes or failures in developing each of these CMVS building blocks will affect the overall system performance. 

The flow chart of the color acquisition procedure is given in Figure 3. The color machine vision system should be white balanced before taking measurement of samples. This can be accomplished by first measuring a white reference target to establish a base line. The R,G,B values of a sample are then acquired. To compare the machine vision measured tri-stimulus values against that of published standard color plates, R,G,B values from CMVS were transformed to L a b values. This transformation required two steps first, R,G,B is converted to X,Y,Z by Eq. 3. Next, Eq. 4 is used to convert X,Y,Z to L, a, b color coordinates. 

Figure 4. The color measurement stability of the color machine vision system. The CMVS has a range of color resolution in measuring various color due to various level of noise in measuring individual colors. The graph shows the magnitude of measurement deviation of each color from 100 repeated measurements.

Figure 5. The color machine vision system and the colorimeter have an R value better than 0.98, in determining color differences between sample colors and a white standard.

Correlation was established between the colors of food samples measured by the colorimeter color and the color machine vision system. The correlation coefficients between L a b color parameters for beef and carrots obtained from the two instruments can be seen in Table 2. For all the data collected, a high correlation was observed (R > = 0.98). The established correlation makes it possible to verify CMVS s color measurement accuracy of colors that are not available from published color standards. Furthermore, the correlations between the CMVS and colorants in food systems can also be inferenced based on established correlations between colorimeter measurements and colorant contents. 

Figure 6. A distribution of the color machine vision system measured colors in Lab color system. There are seven major groups with major color differences. Each group consists of six or seven colors with minor color differences.

Figure 8. Excellent agreement was found between the color machine vision and the colorimeter measured carrot puree colors. The largest disagreement was less than 2 CIE units, while most differences are less than 0.5 CIE unit.

In-line/on-line feedback control of color of food during processing can improve not only color quality but also color related quality such as texture and appearance. To do this, there are three major aspects development of an in-line/on-line color sensor understanding of color change kinetics and establish correlations between instrumental measured and sensory panel perceived colors of foods. In this research, we have chosen color machine vision technology for the measurement of colors of food due to its superior spatial resolution over conventional instruments such as colorimeter or spectrophotometer. Relationships between measured colors and corresponding principal chemical markers were established for the model food systems. We have also found excellent correlations between the color machine vision system (CMVS) measured and a sensory panel determined colors of food samples (Ling and Tepper, 1995). We believed that a CMVS can be used for food process control to ensure color quality as perceived by consumers. 

---