Somatic embryo development is not synchronized. Embryos produced on development medium vary in shape, size and morphology. Separating the high quality (zygotic-like)
embryos from non-embryo tissue and embryos of lower quality by hand is very labor-intensive, and automation is essential to render these steps economically feasible. Machine vision and image analysis systems are being developed currently to automatically count, sort, size and grade somatic embryos of several herbaceous plants such as carrot and sweet potato. Harrell and Cantiffe19 developed a noninvasive machine vision system that could automatically measure the quality of sweet potato somatic embryos produced in an airlift bioreactor. A computer vision system has also been developed for the classification of Norway spruce somatic embryos using algorithm recognition.20 The somatic embryos from development plates were moved with constant speed, and embryos were recognized as they passed a line-scan camera. We have research in progress for sorting of somatic embryos of Douglas fir and loblolly pine using image analysis.21
Fig. 29.3. One replicate plot of zygotic seedlings (from seeds as control) of Douglas fir growing in the field. Photo shows the lesser degree of uniformity as compared to somatic seedlings (Fig. 29.2).
Was this article helpful?