Ozone-mist sterilisation and a web-based sensing system for greenhouse agriculture were studied. Ozone of high concentration of around 70 g/m3 was generated by surface dielectric barrier discharges with high-frequency power. The ozone was injected into water-mist to form ozone-mist matter in a closed glass chamber. Ozone and its mist treatments were performed in a closed chamber (400 mm in height × 500 mm in dia.) which is a model of the greenhouse. The sterilisation rate of pests (aphids) was investigated using the ozone-mist as well as gaseous ozone and chemical synthetics. The ozone-mist sterilisation rate of 90-100% was attained, which is comparable to that achieved by chemicals. A web-based management system combined with the analysing technology of the digital camera images of plant state was developed to predict plant growth and pest development in the greenhouse. It is shown that three colour channels (Red, Green, Blue: RGB) and the greenindex defined by the colour channels are useful to monitor plant growth on a website. Several vegetables were planted in the greenhouse (W: 3 m ×D: 4 m × H: 2 m) and observed by taking digital camera images during 63 days. The daily data of RGB and greenindex were immediately added by sending the real-time images to the remote site through the Internet. It is shown that the greenindex (2 G-R-B) gradually decreases from 42 days to 53 days after the saturated high value during 29 days and 42 days. A web-based greenhouse management system using the digital camera images has the potential to sense plant state in a greenhouse and offers an inexpensive means for small-scale rural farms.