Yield prediction

There are many factors through which a farmer can get optimum results in agriculture. One of these factors is to predict the yield of crop. This factor includes the fertility of soil, irrigation process, climate conditions and controlling of pests. If the farmer does not follow these four factors correctly during farming, there is a huge risk of damaging of crop. Let us see some of the machine learning models which are used in the agriculture sector.

• Machine learning application helps to count the number of coffee seeds in a branch and also it segregates the coffee fruits in three categories of harvest, non-​harvest and seeds with disregarded maturation stage. Also, we can estimate the weight of seeds and maturation percentage of coffee seeds. Ramos et al. (2017) showed how to count coffee fruits automatically from coffee tree using machine vision system (MVS). When the development of crop was going on, during the initial stage and when the harvesting was not done, using MVS technique they proved that estimation of seed count will not be too high or too low and by this they had shown that it obtains higher correlation value of 0.90.
• Amatya et al. (2016) developed an MVS which automatically shakes the trees and catches the cherry fruits during harvesting stage and it also detects the occluded branches and cherries which are not clearly visible. In contrast, during cherry harvesting more labour is required, which take around 50% of its annual production cost. To reduce this cost, mechanized harvesting technologies have been used such as limb-​actuators that vibrate the cherry fruits so that they can be released from the branches.

This tool has generated a new era in the horticulture sector, as it has higher efficiency and is economical for the farmers. Farmers can use this technique in their agricultural work to increase the productivity.