![K-Means Clustering is a method of cluster analysis which
aims to partition n observations into k mutually exclusive
clusters.
Step1: Read Image
Step2: Convert Image from RGB Color Space to
L*a*b*conversion enables to quantify the visual
differences present in the RGB image.
Step 3: Finds partitions such that objects within each
cluster are as close to each other as possible, and
as far from objects in other clusters as possible.
The output of k-means is the set [cluster_index
cluster_center].](https://image.slidesharecdn.com/imageprocessing-150128030527-conversion-gate01/85/Imageprocessing-11-320.jpg)
Imageprocessing
- 2. Plant disease can be indicated by a change in the plant’s appearance relative to a healthy plant of the same age and variety. Healthy plant’s leaf Disease plant’s leaf
- 3. It cause periodic outbreak of diseases which leads to large scale death and famine. In 2007 plant disease losses in Georgia (USA) is approximately $653.06 million (Jean, 2009). In India no estimation has been made but it is more than USA
- 4. Automatic detection of plant diseases is an important research topic as it may prove benefits in monitoring large fields of crops, and thus automatically detect the diseases from the symptoms that appear on the plant leaves.
- 5. Bacterial disease Fungal diseases Viral disease
- 6. Start Capture Diseased Leaf Image Image resize Image Filtering Color Image Segmentation Disease Spots Being Extracted C C Calculate Total Disease Area(Ad) Covert To B/W image Calculate Total Leaf Area(Al) Percent Infection PI=(Ad/Al)*100 Disease Grading By Fuzzy Logic Disease Grade Stop
- 7. The digitization and storage of an image is referred as the image acquisition. In this Process we capture the images of disease leafs. Then all the images are saved in the JPEG format.
- 8. Preprocessing uses the techniques such as image resize, filtering, segmentation, cropping, contrast enhancement, angle correction, morphological operations etc. Initially, captured images are resized to a fixed resolution so as to utilize the storage capacity or to reduce the computational burden in the later processing.
- 9. Noise would disturb the segmentation and the feature extraction of disease spots. So they must be removed or weakened before any further image analysis by applying an appropriate image filtering operation(Gaussian filter ). Resized image Filtered image
- 10. In this process partitioning the digital image into its constituent regions or objects Segmentation should stop when the objects of interest in an application have been isolated . In this system the purpose of segmentation is to identify regions in the image that are likely to qualify as diseased regions. There are various techniques for image segmentation such as clustering methods, compression-based methods, histogram-based methods, region growing methods etc.
- 11. K-Means Clustering is a method of cluster analysis which aims to partition n observations into k mutually exclusive clusters. Step1: Read Image Step2: Convert Image from RGB Color Space to L*a*b*conversion enables to quantify the visual differences present in the RGB image. Step 3: Finds partitions such that objects within each cluster are as close to each other as possible, and as far from objects in other clusters as possible. The output of k-means is the set [cluster_index cluster_center].
- 12. Step 4: Label Every Pixel in the Image Using the Results from K-means with its cluster_index. Step 5: This step will result in k number of images each of which is a segment of the original image that are partitioned by color. When the segmentation is completed, one of the clusters contains the diseased spots being extracted. This image is saved and considered for calculating AD Disease portion is extracted
- 13. In image processing terminology area of a binary image is the total number of on pixels in the image Hence, the original resized image is converted to binary image . From this image total leaf area (AT) is calculated.
- 14. Similarly, the output image from color image segmentation, containing the disease spots, is used to calculate total disease area (AD). Total Area of the original resized image (AT) Total Area of the disease spots image (AD)
- 15. Original Resized Image Disease spots from colour image segmentation
- 16. A Fuzzy Inference System (FIS) is developed for disease grading by referring to the disease scoring scale. Input variable is Percent Infection and output variable is Grade. Percent-infection is given by PI= (AD / AT) *100
- 17. Un trained users initially try to take photos of leaves with multiple leaves present amid clutter. We need a photograph of the leaves against a light, un textured background to prevent the segmentation is challenging due to shadows , bluer, fine scale structure on leaf and specula reflection Leaves vary greatly in shape. This gives rise to complex segmentation boundaries that are difficult to handle for edge-based methods, or region-based methods
- 18. From the result we can observed that the accurate values of percent-infection, and disease grade A proper treatment advisory can be given thereby we can eliminate the above problem