Venomous snake classification by computer vision

Abstract

The identification of venomous snake species for the treatment of snakebite victims is crucial in clinical practice. One approach for diagnosis involves visually inspecting the snake that caused the bite. However, this method can be challenging and prone to errors. To address this issue, author has developed a tool to classify venomous snakes based on images, utilizing computer vision and machine learning technology. Computer vision utilizes the intelligence of computers to recognize and differentiate objects or photographs. This information is then used to diagnose the type of venomous snake. During the training process, images of seven common venomous snake species in Thailand, including Cobras, King cobras, Band krait, Malayan krait, Malayan-pit viper, Green-pit viper, and Russel viper were used. A total of 154 images were utilized to train the computer model. This training enabled the computer to accurately recognize and differentiate images of various venomous snake species with up to 80% accuracy. This tool proves to be highly efficient in classifying venomous snake species based on images. Furthermore, the developed tool can be applied in real-life situations. It allows quick and precise diagnosis by capturing images of the snake and using the developed program. This efficient method significantly enhances the effectiveness of snakebite treatments. This application stands as a prime example of employing cutting-edge technology to solve medical issues, facilitating healthcare professionals in diagnosing complex diseases with ease and precision.

References

อรรถวิทย์ วัชรธรรมรักษ์. โรคงูพิษกัด (Snake bite envenoming) ใน สรุปรายงานการเฝ้าระวังโรค จากการ ประกอบอาชีพและสิ่งแวดล้อม โรคไม่ติดต่อ และโรคจาก การป้องกันการบาดเจ็บ ประจำปี 2562 [Internet]. 2562 [cited 2023 Sep 30]. Available from: https://apps-doe. moph.go.th/boeeng/download/MIX_AW_2_AESR_ 6410-12.pdf

ศูนย์พิษวิทยารามาธิบดี คณะแพทยศาสตร์โรงพยาบาล รามาธิบดี มหาวิทยาลัยมหิดล. Common poisoning: งูพิษ [Internet]. [cited 2023 Sep 28]; Available from: https://www.rama.mahidol.ac.th/poisoncenter/th/ pois-cov/snake

Microsoft. Analyze images with the Computer Vision service [Internet]. [cited 2023 Sep 28]; Available from: https://learn.microsoft.com/en-us/training/ modules/analyze-images-computer-vision

Microsoft. Characteristics and Limitations of Custom Vision [Internet]. [cited 2023 Sep 28]; Available from: https://learn.microsoft.com/en-us/legal/cognitiveservices/custom-vision/custom-vision-cvs-characteris tics-and-limitations

Google. Cloud vision API [Internet]. [cited 2023 Sep 29]; Available from: https://cloud.google.com/vision

สิทธิณัฐ วัฒนมงคล. หนังสือภาพ 7 อสรพิษแห่งประเทศไทย. คราม; 2559.

สวนงู สถานเสาวภา สภากาชาดไทย.ความรู้เบื้องต้นเกี่ยวกับงู. 2554.

image search. https://www.google.com/imghp

A James. Snake classification from images. PeerJ Preprints 2017; 5:1–15.

Patel A, Cheung L, Khatod N, Matijosaitiene I, Arteaga A, Gilkey JW. Revealing the unknown: Real-time recognition of Galápagos snake species using deep learning. Animals 2020;10(5).

Rajabizadeh M, Rezghi M. A comparative study on image-based snake identification using machine learning. Sci Rep 2021;11(1).

Zhang J, Chen X, Song A, Li X. Artificial intelligence -based snakebite identification using snake images, snakebite wound images, and other modalities of information: A systematic review. Int J Med Inform2023;173.