Computer vision is primarily applied as a surveillance system. It can be employed to identify defaulters, set SOPs for working units, thus preventing on-site accidents via alarm systems
Humans have the gift of sight, aiding our daily functions. Everyday, we process images, create memories and identify similar future experiences. Now, Artificial Intelligence (AI), Computer Vision (CV), can help us mirror this gift. However, CV’s visual inputs are much higher in quantity, interpreted at faster rates and yield more accurate outputs. Computer vision is primarily applied as a surveillance system. It can be employed to identify defaulters, set SOPs for working units, thus preventing on-site accidents via alarm systems. Agriculture is the novel experiment zone gaining momentum of CV’s application.
Today about 58% of Indians depend on agriculture for a livelihood. According to IBEF, food grain production hit a new high of 296.65 million tonnes (MT) in the 2019-20 crop year. Moreover, the Indian government envisions 298 MT of food grain production in the current crop year. With more than half of the nation dependent on agribusiness, it is imperative that all processes work in the farmer’s favour. From automating certain functions to detecting discourses overlooked by the naked eye, CV holds the potential to pave the way for smart agriculture.
The promise of a healthy harvest
Traditional crop diagnostic methods depend on a knowledgeable few. The results and eventual solutions, however, might be limiting as the farmers may fail to understand some underlying issues. Their assessments, in some cases, may be misinformed. CV-AI models are more scalable than these traditional assessment methods, further reducing misjudgments, thus increasing chances of a healthy harvest. The data collected can be stored as digitised records for future reference too. The agri-space can benefit from CV models especially for quality control and overall monitoring.
The pre-harvest stage
To detect unfavorable conditions at the initial stages, CV systems can be combined with drones. The aerial data collected can help analyse soil health. The same method can be employed once the crop is in the growth phase. These systems will be able to detect areas that require specific attention. At the micro-level, a closer look at crops shall help detect diseases, pest infestations, or requirements ofq external nutrients. Presently, some smartphone-assisted crop diagnosis applications exist which allow farmers to click images of their crops. It detects issues and provides real-time suggestions for crop care. Once we have the smart analysis, the required medication/solution can be disseminated in the right amount, at the right place and at the right time.
The post-harvest stage
The regular pre-harvest checks, on one hand, cater to reduced crop damage during crop production. On the other hand, post-harvest checks enable the farmers to batch their produce effectively. CV mechanisms checks for width, size and evenness of the batch. It also checks for colorations, moisture and nitrogen levels, depicting the crop’s longevity. The farmers, hence, make informed decisions on the batch to be used immediately, stored for a particular time, or to be sent unharmed to farther locations. Though such systems are dependable, some barriers are yet to be crossed in terms of accessibility.
Ground-zero takeaways
The on-ground reality underscores three key areas of work. Indian farmers mostly have small landholdings, limited accessibility to tech and technophobia. A huge pool of farmers lack the capital, trust and information to utilise the technology. To have an impactful intervention of tech in agriculture, we require well-structured plans of action. It is crucial to raise awareness about the operations, build trust in the farmers and make tech deployment easy. A major step to gain the trust of India’s geographically and vernacularly diverse farmers could be the intertwining of human-tech hybrid models.
Conclusion
As per the Economic Survey 2020-2021, agriculture’s share in GDP reached almost 20% for the first time in the last 17 years. Although the low digital penetration is a reality to reckon with before active adoption, Computer Vision can accelerate the growth of agri-space. Currently, CVs intervention is open to interpretation and ever-evolving. The traditional process is replaced by a time and cost-effective solution. In the long run, it shall enable farmers to make Intelligent agri-business decisions, gain effective ROIs and earn sustainable incomes.
Source: Business World
