Temporal Mango fruit Dataset

The dataset consists of images of mango trees of four different cultivars (Mallika, Raspuri, Arka Arun, and Malgoa), obtained with the main camera of a mobile phone (Oneplus A5000, dual camera (16 MP, f/1.7, 24mm wide, 20 MP, f/2.6, 36mm standard) under natural daylight conditions with an image resolution of 1920*1080 pixels. Data was collected for 5 months temporally in the interval of 7 days for the first 3 months. Then the frequency of data collection is increased for the subsequent two months based on fruit maturity. 800 images per day are captured.
The images were captured in two sessions, namely morning and evening, which is aligned with the direction of the sunlight. Further, we have captured these trees from distances of 2 and 3 meters, respectively.
The selected trees from the orchid vary from 15-20 years of age. The trees range between 7-8 feet high from the
surface. The size of each acquired image ranges between 7-8 MB. The dataset accounts for 21000 images altogether.
In a day, the dataset is collected in two sessions: session 1 was between 9:30 am and 10:00 am and session 2 between
2:30 pm and 3:00 pm from north-east and south-west directions, respectively.
Therefore, the acquired images cover the contour of a tree. However, top and bottom views are exempted from the acquisition. The
dataset was collected throughout a season for five months.