Image Tracking for the High Similarity Drug Tablets Based on Light Intensity Reflective Energy and Artificial Neural Network

Table 2

The reflective energy surface properties of representative templates on the 35th image frame, which were employed as the input vector of ANN for pattern recognition, the same as the other surface property groups.

	The input values of the surface properties
	Property 1	Property 2	Property 3	Property 4	Property 5	Property 6
1	3587.23	548.52	1587.9	10155.4	956.6	0.2669
2	1035.69	669.52	2036.5	12336.5	875.2	0.3022
3	2598.32	716.55	3544.5	10269.5	665.2	0.3654
4	4782.61	215.84	3055.8	9854.6	365.4	0.1485
5	5548.32	985.36	6248.2	9965.5	302.1	0.3954
6	3699.14	888.36	5598.7	7825.5	954.6	0.3022
7	6685.26	441.56	6694.8	6359.4	865.7	0.2035
8	1334.69	332.58	6025.8	9806.5	1012.5	0.3144
9	4026.54	698.22	4785.9	11025.4	1124.5	0.6022
10	4026.57	725.36	6698.5	12004.5	1325.9	0.3598
11	9833.21	889.65	4755.2	7985.6	968.8	0.4752
12	5476.32	102.54	6321.5	9586.6	1024.5	0.3301