QuickBird high spatial resolution images of the maize crop within the study area were acquired. The QuickBird satellite is a high spatial resolution satellite comprising four multi spectral bands (blue, green, red and near infrared) of 2.4 m spatial resolution. The QuickBird satellite image of maize fields was acquired at 09:13 h GMT on 29 June 2007. The image was radiometrically corrected by the supplier of the image (Infoterra group based in the UK). The images were geo-corrected using an image to image technique using ground control points (GCP) collected for fixed points during various field visits covering the entire study site. The images were atmospherically corrected using the FLAASH (Fast Line-of-sight Atmospheric Analysis of Hypercubes) module technique in ENVI v4.9. FLAASH is considered to be more accurate than non-physics-based models such as QUAC (QUick Atmospheric Correction) and is a remarkably established atmospheric compensation algorithm and it is mainly supportive for the majority of multi and hyper spectral remote-sensing platforms. Images were also classified using both unsupervised classification (k-means) and supervised classification (MLC; Maximum Likelihood Classification) to identify different crops in each image. K-means unsupervised and maximum likelihood supervised algorithms were used to identify different crops in the study site. K-means is one of the commonly used unsupervised classification algorithms and during the calculation of overall classification efficiency and Kappa coefficient. We tried both k-means and iso data algorithms and k-means produced higher efficiencies. These two algorithms were performed on QuickBird using the ENVI v4.9 package. The advantage of the k-means is that no additional ground reference points are required to perform the classification. Unlike the k-means technique, MLC requires a reference dataset during imagery acquisition to perform this algorithm. A confusion matrix was derived for both k-means and MLC classifications of the QuickBird satellite imagery. To perform the supervised algorithm, a validation dataset, which was independent from the training dataset, was created manually. The validation dataset included at least 2000 pixels for each class to avoid interference between various classes.

Note: The content above has been extracted from a research article, so it may not display correctly.



Q&A
Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.



We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.