Computer Vision Approach for Outdoor Color Rating of Guava: Design and Implementation

Now a days farmers identifies the ripeness of the fruits by its colors and segregates it by manual judgment. This system is more person orientation and different person may judge it differentially. In the current paper computer-based vision technique is used to find solution to the current problem. Computer based technique is the attempt for achieving automatic and objective based technique developed for reducing the farmer's effort. Computer program to made to make it easier for fruit farmers to rate the color of their fruits. This program can also get rid of shiny spots on the fruits. We tested it in real fields with different lighting, and it worked well. This shows that we can use computers to give fair and accurate color ratings to fruits like guavas, right in the orchard, under natural sunlight. The test was performed on outdoor condition, the result found was satisfactory. We have involved some of the farmer also in the process. They were satisfied with the results. The process for the finding result is very easy and can be done by any illiterate farmer. The initial devices and cost required for the test is quite less. We have conducted test on the 100 samples and got very good results. We have done three set of readings for checking the reproducibility of the program. The test outcomes also indicated that the system's overall accuracy stood at 93%, and we gathered 100 measurements across three distinct outdoor lighting conditions to arrive at this conclusion. Accuracy of the system was 91.99% when Radial basis function. Accuracy of the system was 94% when Linear basis function. This initial research confirmed that it's both possible and practical to utilize computer vision technology for delivering precise ratings. of Guava color in an orchard environment.