Hi Matteo,

I believe it depends on the precision you look forward to reach.

What I personnaly do is place the leaf on a red plank (rigid surface) and take a picture. With a R script (raster package for example), it might be pretty fast.

There is an excellent paper comparing several methods...in French though.

Look here : http://www.oliviertregoat.com/etude_comparative_de_la_precision.pdf

Cheers

Benjamin

I would say it depends on the precision of the leaf area you want.

If you want to describe the grapevine plot scale, I would recommend LIDAR methods (I am currently developing a LIDAR based measuring system for fruit trees) but the resolution is not able to perfectly describe a single leaf.

Benjamin's method is quite good for single leaf area measurement, and you can also use a LI-COR LAI Analyzer for non destructive measurements.

In the end it is a question of scale, and budget!

Thank you both Alexander and Benjamin, very useful comments.

I am in the process of buying some instrumentation to measure LAI and I want to make sure I get a method that is suitable for my needs. It is down to cost in the end, as some of the LAI instrument can be quite expensive, that is why I would like to explore other systems for measuring LAI that do not involve the purchase of expensive instrumentation.

Then I think the photographic option would be the cheapest since you do not need a high-end camera. A red or white (with a little overexposed picture) backgroung is needed, either at the leaf or vinestock scale. Make sure you have a length reference on the picture, and that your picture compression is 1:1 pixel (otherwise you have to know the pixel ratio in order to convert area in pixels to the proper area in cm2 or in2).

I personaly use ImageJ to process this kind of pictures.

I guess Benjamin would know a lot better than I do about grapevine specific studies.

Thanks a lot Alexandre, I have used ImageJ in the past for image processing (not for leaves) so I think I will give it a go

Cheers

I'd follow the wise recommandations from Alexandre concerning the image analysis (size, scale, definitions). Have fun!

when very sofisticated insytruments were not available I sampled a large number of leaves of different size and measure the length and rhe width, than by a leafmeter I measure the surface, than i calculated the correlation between the measures and the leaf area. The best correlation for Sangiovese grapevine was beteew higthness x widenes, although the heigth of the leaf gave a good correlation. So the way was measure the length of of the main veins in the field of several leaves and calculate dby the correlation the area. On Cabernet Sauvignon which has an horbicolar leaf shape the correlation was lesser even though measure the width was not necessary.

of course in this case one must go in the field to take measumerents, and the correlation must be done for each variety.

It is possible also take less measuremets in the field if one is able to take measumements of leaves of different positions, so as one could obtain a curve of leaf area according to the ength of tjhe shoot and the leaf number. In this case the number of leaves to take to built a database is higher as laterals are bearing lesser and smaller leaves whic do not follow a dinamic area like those of the primary shoot.

Hi, Iam using image j to measure leaf area but i think it is not accurate with all my leaves especially i have compound leaf and I am looking for another way to measure leaf area from length and width . Any recommendation from users would be welcome.