Citrus Leaf Disease Detection
Agriculture is a vital component of Pakistan's economy as it provides the raw materials for industry down the line and helps alleviate poverty. This sector has contributed 19.8 per cent of GDP and remains by far the largest employer to absorb 42.3 per cent of the total labor force in the country. Cu
2025-06-28 16:30:47 - Adil Khan
Citrus Leaf Disease Detection
Project Area of Specialization Artificial IntelligenceProject SummaryAgriculture is a vital component of Pakistan's economy as it provides the raw materials for industry down the line and helps alleviate poverty. This sector has contributed 19.8 per cent of GDP and remains by far the largest employer to absorb 42.3 per cent of the total labor force in the country. Currently, citrus fruits are grown in Pakistan on an area of 206,569 hectares with a production of 2.36 million tons in 2015-16. Citrus fruits are divided into different groups of sweet oranges, mandarin, grapefruit, lemon and lime which are commercially grown. The agriculture sector of Pakistan has been producing vast amount of citrus fruits. Considerable amount of Kino is produced throughout the entire region of Punjab and it supplies 2.3 million Kino per year. The agricultural institutes states the loss of citrus production have increased in the last few years, due to the spread of citrus diseases, causing fruit and in some cases whole tree to get infected, making it a non-consumable product. More than 98% of citrus fruit is produced in Punjab and 70% of citrus fruit is produced in Kino. In fact, in the country, Kino monopolized citrus farming. Unmaintained fields of citrus fruit result in rotten and useless crops. Not every farmer has proper knowledge of the disease that citrus fruit can process (especially in early stages). This results in great loss of economic resources of the country. Providing services to every field to check whether the crops are suffering from disease, which makes managing the fields difficult. Citrus fruits are infected through the leaves which are contaminated due to diseases like citrus canker and citrus greening. Escalation of such diseases leads to the ruin of citrus fields which can cause farmers to lose a considerable amount of income. Early detection of such diseases can save citrus fields which will, in turn, lead to effective growth in Pakistan’s GDP. As the agriculture sector contributes 42% in the entire GDP. For the detection of the diseases, we will use a rear RGB camera of an android cellphone. The image taken from a camera will be subjected to image reprocessing. Features will then be extracted from the image and through these features, machine learning algorithms like SVM, K-MEANS and CNN will be able to classify and detect the disease.
Project ObjectivesWe will develop an android application to detect citrus leaf diseases using android phone captured images. Target audience for this project are the farmers having android phones. The objectives for this project are listed below:
- Data set collection
- Data set annotation
- Training model
- Disease detection
- Achieving high accuracy
- Providing user interface
Data set will be collected through field visits. The data collected from the field will then be annotated. Using image preprocessing techniques features will be extracted from the image and then a model trained using the collected data set will classify the disease.
Benefits of the ProjectBy using this system, the agriculture sector of Pakistan will be able to use the disease detection facility free of cost providing them with the benefit of free disease detection software running on their smartphones. This will enable huge amount of farmers to detect the disease and cure it before it infects the whole tree. Hence saving the large quantity of citrus fruits. This will gradually increase the profit of the farmers and will also have a positive effect on the GDP of Pakistan, as Pakistan’s entire GDP is comprised of 24% of agriculture sector, which produces approximately 2.3 million tons in 2016-16.
Technical Details of Final DeliverableOur system, an android application, will be able to detect diseases on the citrus leaf using methods of machine learning and image processing. Our system will detect the disease on the citrus leaf. The user will capture the image of the leaf using their android phone. Then application will use image pre-processing techniques
Final Deliverable of the Project HW/SW integrated systemCore Industry AgricultureOther IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Zero HungerRequired Resources| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 80000 | |||
| Field visit for Data set collection | Miscellaneous | 4 | 2500 | 10000 |
| Ram 8gb ddr3 | Equipment | 1 | 4000 | 4000 |
| Power Adapter | Equipment | 2 | 1000 | 2000 |
| Memcard- 32GB | Equipment | 2 | 1200 | 2400 |
| Tool kit | Equipment | 3 | 1000 | 3000 |
| Wires AWG set | Equipment | 3 | 2000 | 6000 |
| Capacitors | Equipment | 20 | 30 | 600 |
| connectors | Equipment | 5 | 1500 | 7500 |
| raspberry pi 3 model B | Equipment | 2 | 6500 | 13000 |
| Arduino | Equipment | 5 | 1600 | 8000 |
| Battery pack | Equipment | 1 | 20000 | 20000 |
| Esp 8266 | Equipment | 1 | 1500 | 1500 |
| Clamp meter | Equipment | 1 | 2000 | 2000 |