An image analysis system is described to study the process of imbibition in white cabbage (Brassica oleracea L.) seeds. The system consists of two components: 1) an environmental system where seeds germinate; 2) a computer imaging system, composed of a Charged Coupled Device (CCD)-imaging sensor, an image frame grabber, a computer and a video monitor. The measurement software can determine area, perimeter, width, length and eccentricity as descriptors of changes in seed size during swelling. The captured images of imbibing seeds allowed estimation of timing of 'visible germination' occurrence without any subjective evaluation. Using the area parameter, the imbibition process can be described graphically by a series of curves similar to the triphasic pattern of water uptake, with extension and rate depending on the degree of seed viability and germination medium conditions. Imbibitional osmotic stress, provided by -1.5 MPa mannitol, induced a large delay in germination and, upon stress removal, a more uniform radicle emergence. The versatility and the sensitivity of the method, together with the feasibility to investigate germination rate in individual seeds within a population, suggest that image analysis techniques have high potential in seed biology studies.

Internal cracks caused by high temperature or excessive moisture during maize (Zea mays L.) kernel development were characterized, and their effects on kernel quality were assessed. Pre-harvest stress cracks are often located near the middle of the kernel along the embryo axis, but they were also detected in other positions, irrespective of the shape of the kernel. X-ray analysis enabled visualisation of stress cracks that are invisible to the human eye and, therefore, gave a better estimate of the percentage of cracks. However, low temperature scanning electron microscopy of the surface of milled kernels revealed small cracks not noticed by visual or X-ray inspection. All kernels tested in this way had a crack of some sort in the endosperm tissue. Cracks were also frequent in the scutellum, but rare in the embryo axis. Endosperm cracks followed the boundary of the starch granules, but did not extend into the pericarp tissue. In contrast to external cracks caused by mechanical impact, preharvest internal stress cracks generally are not detrimental to germination and vigour. However, if the crack is located inside or perpendicular to the embryo axis, it may affect the quality of the kernel, probably by impeding nutrient translocation to the embryo.

An accurate method was developed, using image analysis, for measurement of the length width and shape of cucumber fruits. Determination of the local width along the mid-line of the fruit, enabled a condensed description of the fruit to be given. From this condensed description, size and shape features can be extracted, such as length, width and neck-shape of cucumber fruits. Tests showed that the image analysis method was effective and efficient for determining shape features which are visually difficult to score. Comparison with the manually obtained results showed that the measurement errors for image analysis were much lower than the ones made by the observer, e.g. the standard deviation of cucumber length was 1·4 mm with image analysis and 3·5 mm by manual measurement. The proposed method has more general applicability and may be of interest for studying the shape of other elongated symmetrical objects.