If one tries to build perception as dependent on legacies and/or cultivations of values either in the cultural or the religious sense, then the metaphorics of judgments hovering around values locks away the empirical investigations. This is beautifully outlined in the work of Wilfred Sellars‘ Science, Perception and Reality. For Sellars, perception is a non-inferential knowledge. But, it has epistemic significance, since it is gathered via the labors of the the way language is learned and acquired. Further on, his distinction between the manifest and the scientific image gives legitimacy to the issue with the former being the commonsensical conceptual template that enables the practice of philosophy in the first place, since through this, man gets to be aware of his existence as man-in-the-world. This is the point where Sellars shows his allegiance to the Platonic-Aristotelian western philosophical framework. In contrast to the manifest image, scientific image drives on by using postulated entities at its core, thus obligating the not-so-necessary inclusion of the awareness of man-in-the-world. Perception is built heavily into the manifest image in comparison to the scientific one.
On to a distinction between perception and religion, I cannot say much excepting the maintenance of a water-tight distinction in order to preserve the normative kernel of rationalism and trading it off as regards the metaphysical shell. Perception is vulnerable to fall into the latter, and therefore in the process is likely to hurt Sellarsian and Brandomian clinging on to the historicity of reason in the true Hegelian sense, and subsequently elevate the neo-Aristotelian substantialization of mind which mires itself in Hegelian theological orthodoxy. I emphasize this, since I suggest Sellars as the point de capiton. I still maintain Sellars and even for that matter the present day analytical philosopher Robert Brandom to be copiously converging towards the thesis of perception versus religion. One way to create this distinction would be call back inferentialism (Brandom does it perfect!). Inferentialism sticks on to methodological naturalism, and thus creates spaces for folk-psychological theorizing. The best person that comes to my mind in this regard happens to be an American anthropologist Pascal Goyer, who in his Religion Explained unequivocally sets out to promulgate the mentally constructive and perceptual evolution of the mind-brain based on a much better understanding of critiques of religious representation.