Even before the pandemic, there were signs that global food prices could

soon surge. Extreme weather events induced by climate change have

become more common. African swine fever wiped out over one-quarter of the
world’s pig population last year, causing food prices in China to increase by
15-22% year-on-year so far in 2020. And, more recently, the worst locust
blight in 70 years has destroyed crops in East Africa. In Kenya, the price of
maize, a staple food, has risen by over 60% since 2019.

COVID-19 is amplifying the risk of a worldwide food-price spike, which would

trigger crises in many developing countries. In the poorest of these, food
accounts for 40-60% of the consumption basket, about 5-6 times its share in
advanced economies.

Furthermore, some major food-producing countries have already imposed

export bans or quotas in response to the pandemic, as Russia and
Kazakhstan have done for grain, and India and Vietnam have done for rice.
Meanwhile, other countries are stockpiling food through accelerated imports,
as is true of the Philippines (rice) and Egypt (wheat).

More recently, the corona-virus pandemic triggers the risk of increasing food-price in various countries specially in
developing countries. Moreover, some major crop-producing countries declared not to export in this pandemic
situation. Russia and Kazakhstan give veto for grain, and on the contrary, Egypt imports wheat for further food
security. In the case of rice, India and Vietnam prohibit exporting in the recent future; on the other hand, Philippines
stocks. Before the Corona-virus pandemic, in East Africa, crops have been destroyed due to the attack of the locust
swarm. Further, from 2019, in Kenya, the price of corn (maize), staple food has been alarmingly increased by 60%.
 Figure 7. Some true positive samples of plant disease images: (a 1) is rice sheath
rot image having challenges of sun light and it makes difficult to trace
yellow symptoms with yellow sun light shades and (a 2) represents the
located disease area of this leaf image.(b 1) is rice blast disease image in
complex background because of same color of disease region and
background, and (b2) represents the located disease area of this leaf
image. (c1) is rice blast image with natural background and having
disease region in the upper part of the shrunk leaf image and (c2)
represents the located disease area of this leaf image. (d 1) is rice
leafblight rotated image with overlapped leaf and blur light, and (d2)
represents the located disease area of this leaf image. (e 1) is shrunk rice
leafblight image and (e2) represents the located disease area of this leaf
image.

Figure 9. Some misclassified images: (a), (b), (c) are some false positive rice
sheath-rot images. (d) is rice bacterial leaf blight.(e) and (f) are some
false positive potato late blight images. (g) is false positive apple
black-rot image and (h) is false positive corn northern leaf-blight.

Figure 8: ROC Curve for Histogram based Gradient Boosting Classifier.

Figure 7. Some true positive samples of plant disease images: (a1) Rice
Sheath-rot image having challenges of sun light and it makes difficult to
trace yellow symptoms with yellow sun light shades, (b) rice blast
disease image in complex background because of same color of disease
Figure 3. Various plant disease images: (a) Corn Northern Leaf-blight,
region and background, (c)
(b)Pepper Bell Bacterial-Spot, (c) Grape Black-measles, (d)
in natural background, (d) rotated image, (e) shrunk
Rice blast,(e) Potato Early-blight, (f) Apple Black-rot, (g)
image and (f),(g),(h),(i) and (j) are their segmented
Mango Sooty-mould, (h)Cherry Powdery-mildew,(i)Rice

Bacterial Leaf-blight,(j) Potato Late-blight, (k) Rice Sheath-

rot and (l) Rice Tungro