Precision in horticulture refers to a management strategy using electronic information amalgamating other technologies to gather, process and analyze spatial and temporal data. The goal of this strategy is to optimize agricultural outputs. Horticulture significantly contributes to Indian economy by augmenting farm output, generating employment and supplying raw materials. In contrast to traditional farming practices, the lack of advanced technologies for soil, light and temperature control, crop monitoring, water management, and pest and disease identification remains a major challenge in horticultural production. This study was conducted in Sonarpur, Mathurapur, Baruipur, Jaynagar and Lakhmikantapur blocks of South 24 Parganas, West Bengal, India. The study focused on the cultivation of Solanum lycopersicum, Abelmoschus esculentus, Solanum melongena, Capsicum annuum, Raphanus sativus, Luffa acutangula, Lagenaria siceraria, Momordica charantica, Cucumis sativus and Cucurbita pepo in mono-culture fields from January to December 2024. Transfer learning using ImageNet has been utilized for identification of vegetables whose pictures are already available. Images of remaining species (Lagenaria siceraria, Capsicum annuum, Abelmoschus esculentus and Cucurbita pepo) were also obtained from high-resolution RGB aerial cameras operating from 30, 40 and 50 meters above ground. Approximately 20 video frames per tree were captured, with a shift of 20 pixels per frame. KNN, SVM and Naïve Bayes were used for classification and detection of pest and diseases of crops also predicting crop losses. Equipped with computer vision, drones could monitor the quality of crop growth and minimize damage. The application of automated precision irrigation could also reduce wastages, improving resource utilization. AlexNet, VGG-16, ResNet- 50, Faster RCNN, YOLO v3, Mask RCNN, Inception ResNet architectural models were utilized for image processing. Mask-RCNN was helpful in detecting and counting the number of fruits while YOLOv3 proved beneficial with fruit localization. Fruit classification based upon the ripening stage was carried out using AlexNet, ResNet, VGG-16 and Inception Net. The ResNet model displayed an accuracy of 86.54%, F1 score of 0.849 and recall score of 84.1% for fruit detection. AlexNet also yielded fruitful results with a recall score of 88.14%, F1 score of 0.854 and an accuracy of 86.75%. Faster RCNN showed greater performance (mean average precision i.e. mAP-83%) as compared to Mask-RCNN (mAP-72.3%) and VGG-16 (mAP-70.38%) for disease identification. SVM classification reached the highest accuracy of 86.4% compared to KNN (81.4%) and Naïve Bayes (76%) for pest detection and classification. Future research could explore the integration of Industry 4.0 technologies, such IoT, cloud computing and blockchain, to further improve horticultural practices, optimize resource consumption and promote sustainability.