Diabetic retinopathy (DR) is an eye condition caused by uncontrolled diabetes and can affect the retinal tissues and lead to vision impairment if left untreated for a long period. There are many computer-aided techniques and artificial intelligence algorithms to detect lesions. Detecting fundus lesions through these techniques is time-consuming, expensive, and more complex even for ophthalmologists. The contrast of image and noise is the challenge during DR diagnosis. Thus, the spot highlighting method and Contrast-Limited Adaptive Histogram Equalization (CLAHE) process are used to enhance the features of the eye fundus images towards detection of fundus lesions for diabetic retinopathy analysis. In addition to spot highlighting and CLAHE, data amalgamation techniques are used to overcome the challenges of over-fitting and noise.