3D PLANT GROWTH MEASUREMENT

3D Plant Growth Measurement



3D Plant Growth Measurement

Introduction

Interferometry is a method of superposing (interfering) two or more swell, which conceives a yield signal distinct from the input waves. This in turn can be utilized to discover the dissimilarities between the input waves. Interferometry procedure is adept to give unquestionable displacement information.

As early as 1976, Fox and Puffer suggested their holographic interferometer set about in their study of vegetation development measurement. Based on some preceding study utilizing holographic interferograms, this Fox and Puffer evolved this procedure farther and did quantitative measurements of transient vegetation movements utilizing this method. They computed the displacement vector of the fact issue on study vegetation from the holographic interferograms. Holographic interferometer presents a enduring, highly unquestionable estimation outcome, as an optical procedure it has no direct communicate with the vegetation, and hologram encompassing whole vegetation likeness makes it likely to quantify displacement, velocity and acceleration at all evident fact issue on plant. They were the first to get dynamic transient vegetation shift facts and numbers over a short time frame. The method consists in subtracting two consecutives 3D surface scan images (essentially two range images in the case of plant growth analysis) taken at different times. The range images are taken directly above the plants to be monitored. The images consist of an array of elevation measurements. Typically the image array has 512 by 512 pixels (i.e. elevation measurements).

Each measurement has some level of noise or uncertainty associated with it. If si is the noise level for individual measurement, then the resulting noise in the differential image will be:

Where ad is the differential image noise level. Typically, range noise levels for triangulation-based optical techniques are in the order of 10 to 100 µm. In the following experiment the noise level was measured at 40 µm. This means that the differential noise level is at about 60 µm. What is making the approach powerful is that more than 100,000 elevation values can be averaged, in a single differential image, to extract a very accurate global growth rate for all of the plants under the viewing field.

The 3D laser digitizer is a synchronized scanner camera [6] directly positioned above the area to be monitored. In this mode of operation a range image is obtained. The acquisition of image data is similar to a flatbed scanner, with the exception here, that each pixel has elevation information about the origin of the scattered laser spot on the vegetation.

The experiment consists in taking two range images at two different times. For this test, we record a first range image, then, we immediately add water to the plants and wait for 26 minutes before taking the second range image. Figure 1 shows the first range image. Grey level coding is used to display elevation, white being closer to the 3D camera than black.

Figure 2 shows the difference between the two range images. Most values are around zero except for extreme cases where leave have substantially moved ...