Plant Species Recognition based in leaf photo
is mobile system , to help professionals and amateurs in plant classification to recognizing and identifying a plant species by taking a photo of its leaf , using common smart phone's or PDA's cameras . In this project we used common computer vision technics for Image processing, and mathematical methods for classification and identification. We have successfully designed, developed and implemented a system that can work efficiently, and achieved a performance of 90.9% accuracy for first guess of plant species, and 99.45% in finding the right specie in first five answers. Our System can work efficiently in various types and levels of smart phone's cameras.
Our System works in nature environment where a leaf is commonly found, i.e. parks, forests, despite of noisiness and effect of shades. System has been implemented as web application, and a smart phone WEB based application In the next step.
here I will make a presentation of the WEB APP !!
the web design tablet & smart phone oriented
the software engineering design of the system and the Ai research
will not be discussed here .
the video will be availabe soon.
tools :
Asp.net
MS SQL server
CSS 3
javaScript ( JQuery mobile )
is mobile system , to help professionals and amateurs in plant classification to recognizing and identifying a plant species by taking a photo of its leaf , using common smart phone's or PDA's cameras . In this project we used common computer vision technics for Image processing, and mathematical methods for classification and identification. We have successfully designed, developed and implemented a system that can work efficiently, and achieved a performance of 90.9% accuracy for first guess of plant species, and 99.45% in finding the right specie in first five answers. Our System can work efficiently in various types and levels of smart phone's cameras.
Our System works in nature environment where a leaf is commonly found, i.e. parks, forests, despite of noisiness and effect of shades. System has been implemented as web application, and a smart phone WEB based application In the next step.
here I will make a presentation of the WEB APP !!
the web design tablet & smart phone oriented
the software engineering design of the system and the Ai research
will not be discussed here .
the video will be availabe soon.
tools :
Asp.net
MS SQL server
CSS 3
javaScript ( JQuery mobile )
Book cover :
Print screen of the web app on ( Samsung galaxy tab 10.1 ) :
the process of the web app .
1- the user choose a image or take a photo .
2- click detect ( to detect the leaf in the photo )
3- click recognize ( to analysis the photo of the leaf and find out what are the spices most similar show them on the right side most similar to least )
4- the user can browse throw the results and click in any one to see more images that this is
the plant in front of him and read the information and this plant .
Picture of the app running on ( Samsung galaxy tab 10.1 ) :
this WEB app designed ( CSS / JavaScript ) to be fully compatible with tablet and smartphone .
it respond to the device rotation :
here you can see the layout design and the states of the interface .
1- the area change so we don't just show the objects smaller but hide some non important objects ( in this stage and show others )
2- !! [ state of use ] instead of add buttons of preferences .. I thought of how the user would use the tablet ( hold it ) to implement the two states
> state 1 : he is taking photos so mostly he will hold the tablet in the landscape mode while taking photos . so the control panel will be shown with 40% width and the remaining width for the results .
> state 2 : when the person want to focus on the results he will rotate the tablet so he will use the width of the screen . here the control panel will be hidden and the images will be much bigger , so he don't need to open the links just to see the images bigger .
thats my point of view , a lot of designers keep thinking of the buttons like the old PC computers instead of thinking of the sensors , cams , GPS ..... etc
Hello ... the machines now a days are much smarter they can see , feel , hear ... and get a direct orders :)
when you the client browse throw photos of spices the script very response to the touch gesture slide , swip , pinch , rotate ... and the performance very high because in build using JQuery mobile .
one last thing about why I prefer web app .
1- no installation / update needed
2- work on different environments ( android , ios .... )
...
but on other hand .
1- it new an internet connection ( cant work offline ! ).
2- traditional mobile app can do a little preprocessing
...
< the video is under maintenance >
