Analysis of damaged area on leaves may be done on images with single or multiple leaves. In images containing multiple leaves, percent damage may be obtained for each individual leaf. Quantifying damaged areas of leaves requires two color separations. The leaves must first be identified from the image background using the procedure in the area analysis protocol. Damaged areas are then identified on the leaves. This protocol outlines key parameters that must be set for IPLab to analyze images correctly. All of the parameters and dialog boxes shown below are in the impatiens damage batch analysis script. The script may be modified for the specific analysis being conducted. See scripting protocol for instructions. A color card was included in the images in this example, although it was cropped out manually before analysis. Alternately, it is possible to exclude the color card from analysis by size parameters or by telling IPLab where the card is located in the image. See IPLab manual for further details.
| Step | Image |
| For the script to run the images must be saved in sequential numerical order in the "working" directory on the C drive of the computer Thrips 3. Open .tiff image in IPLab. Zoom in or out on the image if necessary. The white and brownish areas on the leaves are thrips damage. It is important that the leaves do not have petioles attached and are not touching each other. It is best to have space between the leaves when taking the image, although it is possible to separate leaves with a thin black line in Corel PhotoPaint. | |
| Follow area analysis protocol to identify individual leaves. The parameters at right were used to identify leaves. These may need to be adjusted depending on leaf color, what type of system was used to capture the images, etc. This step along with the next two produce the following image of the leaves in the script. | |
| Several options (Analyze-->Set Measurements) are available to define what you want the software to analyze. The images we work with typically have some spots in the background that are identified along with the leaves. To exclude this noise the "Area" box and the corresponding "Limit?" box are both checked, with appropriate values entered for min and max. | |
| The Analyze-->Measurement Options box further defines what will be analyzed and how the segmentation will be displayed. Segment color is chosen and segments excluded based on the previous step can be erased. We want the area of the entire leaf, so including interior holes is checked. Segments touching the edge of the image can be ignored by checking the appropriate box. Under Annotation, we choose to outline segments and number segments. The numbering is critical for each leaf to be analyzed individually. The Results Window sets the name for the leaf area output and tells IPLab to keep the image name with the data from that image. | |
| At this point in the script, the output at right is generated. The leaves are counted, with the output being placed in the "leafcount" window. | |
| The next step is to identify damaged areas within the leaves. This is the same procedure as the area analysis process. However, we want to find a separation method(s) that allow us to select only the damaged areas. Parameters may then be entered into the script. Hue and Yellow were the components which best identified the damge in this case. Components or the range within the components may need to be adjusted depending on the images. | |
| The Analyze-->Set Measurements box is included in the script for the damage identification, although nothing is selected. | |
| In the Analyze-->Measurement Options box a different segment color is selected, interior holes are included in measurements, and the outline color is changed to yellow. Nothing else is selected. In the Results Window section, the name is damageresults and the image name is used to identify the data associated with the image. | |
| The damaged areas show up as magenta segments on top of the green segments. The "damageresults" window contains the data for this analysis. The total area of damage and the %ROI, or Region Of Interest (in this case each leaf) are given. The impatiens script is set up to leave the original image in the working directory, save the image at right as an IPLab file, and save the "leafcount" and "damageresult" windows as text files which may opened directly into excel. | |