Identification of Tomato Disease Types and Detection of Infected Areas Based on Deep Convolutional Neural Networks and Object Detection Techniques

Table 2

The specific information of the dataset.

Diseases	Original^1	Expanded^2	Training^3	Validation^4	Test^5
Healthy tomato	64	320	150	105	65
Tomato malformed fruit	38	190	90	60	40
Tomato blotchy ripening	16	80	35	25	20
Tomato puffy fruit	22	110	50	35	25
Tomato dehiscent fruit	35	175	80	60	35
Tomato blossom-end rot	18	90	40	30	20
Tomato sunscald	14	70	30	25	15
Tomato virus disease	34	170	80	55	35
Tomato gray mold	28	140	65	45	30
Tomato ulcer disease	9	45	20	15	10
Tomato anthracnose	8	40	15	15	10
Total	286	1430	655	470	305

¹Number of original images; ²number of processed images; ³number of images in the training set; ⁴number of images in the validation set; ⁵number of images in the test set.